Electron- and photon-impact atomic ionization

I. Bray, D. V. Fursa, A. S. Kadyrov, A. T. Stelbovics

ARC Centre for Antimatter-Matter Studies, Curtin University, GPO Box U1987, Perth,
WA 6845, Australia

A. S. Kheifets
Research School of Physical Sciences, The Australian National University, Canberra
ACT 0200, Australia

Abstract
The last decade has seen extraordinary theoretical progress in the field of fully
kinematically resolved atomic electron-impact ionization, and its closely re-
lated field of double photoionization. These processes were challenging to cal-
culate due to formal and computational difficulties associated with breakup
problems involving the long ranged Coulomb potential. Presently, however,
these processes can be routinely calculated accurately for suitable targets,
irrespective of the kinematics considered or the geometry of detectors. We
report on the computational progress, and how it has resulted in a deeper
understanding of the formalism of Coulomb few-body problems.
Keywords: electron-impact ionization, double photoionization

Contents

1 Introduction 2

2 Formal scattering theory for Coulomb few-body problems 3

2.1 Formal problems of the conventional theory of ionisation . . . 7
2.2 Surface-integral approach . . . . . . . . . . . . . . . . . . . . . 12

3 Computational methods for Coulomb breakup problems 25

3.1 Exterior complex scaling . . . . . . . . . . . . . . . . . . . . . 25
3.2 Time-dependent close-coupling . . . . . . . . . . . . . . . . . . 28
3.3 Convergent close-coupling . . . . . . . . . . . . . . . . . . . . 30

Preprint submitted to Physics Reports March 12, 2012

 3.4 R-matrix with pseudo-states . . . . . . . . . . . . . . . . . . . 33
3.5 Integro-differential close-coupling . . . . . . . . . . . . . . . . 36

4 Electron-impact ionization 37
4.1 Hydrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.1.1 S-wave model . . . . . . . . . . . . . . . . . . . . . . . 37
4.1.2 Low energies . . . . . . . . . . . . . . . . . . . . . . . . 40
4.1.3 Intermediate energies . . . . . . . . . . . . . . . . . . . 44
4.1.4 High energies . . . . . . . . . . . . . . . . . . . . . . . 45
4.2 Helium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.2.1 Low energies . . . . . . . . . . . . . . . . . . . . . . . . 51
4.2.2 Intermediate energies . . . . . . . . . . . . . . . . . . . 51
4.2.3 High energies . . . . . . . . . . . . . . . . . . . . . . . 56
4.2.4 Ionization plus excitation . . . . . . . . . . . . . . . . . 58

5 Double photoionization 59
5.1 Helium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
5.1.1 Total integrated cross-sections . . . . . . . . . . . . . . 62
5.1.2 Fully differential cross-sections . . . . . . . . . . . . . . 63
5.1.3 Symmetrized amplitudes . . . . . . . . . . . . . . . . . 63
5.2 Other targets . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
5.2.1 Quasi two-electron targets . . . . . . . . . . . . . . . . 64
5.2.2 Three-electron targets . . . . . . . . . . . . . . . . . . 65
5.3 Other related processes . . . . . . . . . . . . . . . . . . . . . . 66

6 Concluding remarks 66

1. Introduction
Collisions in the realm of atomic physics not only have many practical
applications, but also form the testing ground for the underlying quantum
collision theory, made possible by the wealth of experimental data. Due to
the long-range nature of the Coulomb potential charged particles continue to
interact with each other even at infinite separation. For this reason the much
studied problem of electron-impact ionization, and its very close relative of
non-sequential double photoionization, have lacked a proper formal founda-
tion despite many successful computational implementations. In fact it was

2
the success of the computational approaches to the problems that drove us
to revisit the underlying formal theory.
The purpose of this article is to review the recent progress in formal
ionization theory, and show how it relates to the successful computational
techniques which aim to fully solve the ionization problems without resort-
ing to approximations which limit the applicability of the methods. Such
issues are best addressed using relative simple targets of atomic hydrogen
and helium, where we can be very confident in the accuracy of the calculated
target structure. Additionally, we shall restrict detailed discussion to the
lower energies, where high-energy approximations are inaccurate.
We begin with the discussion of formal issues in Sec. 2 and the computa-
tional methods in Sec. 3. This is followed by applications to electron-impact
ionization in Sec. 4 and double photoionization in Sec. 5.

2. Formal scattering theory for Coulomb few-body problems

Scattering in a three-body system at energies above the three-body breakup
threshold is one of the central subjects of quantum mechanics. Enormous
progress has been achieved in recent years in describing (e,2e) processes via
the exterior complex scaling (ECS) [1–3], convergent close coupling (CCC)
[4–6] , R-matrix with pseudostates (RMPS) [7–10] and time-dependent close-
coupling (TDCC) [11] methods. As described in the forthcoming sections of
the present review, kinematically complete picture of electron impact ioni-
sation (of simple atoms) is well understood, with calculations fully support-
ing experimental observations. The electron-hydrogen breakup problem is
considered as practically solved. In particular, the success of the ECS and
CCC approaches to the fundamental Coulomb breakup problem caused re-
examination of the underlying formal theory [12–14] which culminated in
a surface-integral approach [15] to formulating scattering theory. We start
from a brief review of this approach in a form applicable to arbitrary three-
body systems with both short-range and Coulombic long-range potentials.
The formalism is based on a surface-integral approach. New prior and post
forms of the breakup amplitude for a three-body system will be given that are
valid for both short-range and Coulombic potentials. This resolves a problem
of the missing conventional post form of the Coulomb breakup amplitude.
An essential feature of the surface-integral formulation is that it avoids any
reference to the Green’s functions and formal solutions of the Schrödinger
equation in integral forms. Therefore, for the purpose of defining the scat-

3
tering amplitudes the knowledge of a complicated analytic structure of the
Green’s function in the complex-energy plane is not required. This con-
stitutes a simpler yet more general alternative to formulations adopted in
textbooks on scattering theory.
Microscopic many-particle systems with the long-range Coulomb inter-
actions and cosmological many-body systems with the gravitational interac-
tions are of fundamental importance for our understanding of the universe.
Atomic and nuclear systems are governed by the pairwise Coulomb potential
(in addition to other possible short-range interactions) which is proportional
to the charges of interacting particles and inversely proportional to the in-
terparticle distance (r). In case of cosmic objects the masses of the objects
play the role of gravitational charges. This 1/r potential is arguably the
most important interaction in the structure of the universe. The long-range
character of this potential implies that there is interaction even at very large
(in fact, infinitely large) distances. This leads to various divergency problems
in many branches of physics from quantum mechanics to quantum electro-
dynamics, from field theories to statistical physics, from celestial mechanics
to cosmology. Despite being identified from the very beginning of modern
physics, these problems have resisted proper solution. As a result different
renormalization theories have been developed aimed at their ad hoc resolu-
tion. However, such approaching have been possible only in some limited
class of problems. One can think that a simple cut off beyond some finite
range could be a remedy for these problems. Unfortunately, this is not the
case for most systems of practical interest. This could be because limiting
the range of the potential leads to the loss of information about the very
nature of the field creating the potential that cannot be recovered by simply
taking the screening range to infinity.
It is well known that conventional quantum-mechanical collision theory
is valid only when particles interact via short-range potentials. For charged
particles the theory requires modification due to the fact that the long range
of the Coulomb potential distorts the incident and scattered waves right out
to infinity. Formal scattering theory is generalized to include Coulomb long-
range potentials using renormalization methods [16]. The renormalization
theories lead to the correct cross sections for the two-body problem, how-
ever, the results from these procedures cannot be regarded as completely
consistent. For instance, in screening-based renormalization methods differ-
ent ways of screening lead to different asymptotic forms for the scattering
wave function. These asymptotic forms differ from the exact one obtained

4
from the solution of the Schrödinger equation. Moreover the renormalization
methods give rise to a scattering amplitude that does not exist on the en-
ergy shell. In other words, the resulting amplitude cannot be directly used
for calculating cross sections. This is because the amplitude obtained in
these methods has complex factors which are divergent on the energy shell.
Therefore, these factors must be removed (renormalized) before approaching
the on-shell point. Furthermore, the renormalization factors depend on the
way the limits are taken when the on-shell point is approached. In other
words, depending on the way you take the limits different factors need to be
removed. Thus, the ad-hoc renormalization procedure is not logically con-
sistent. In quantum collision theory it is customary to define the scattering
amplitude in terms of the scattering wavefunction and the potential of inter-
action. Despite the fact that the Coulomb wavefunction and the Coulomb
potential are both known analytically, the conventional theory has not been
able to provide such a standard definition for the amplitude of scattering of
two charged particles, which yields the Rutherford cross section.
The situation with the Coulomb few-body scattering problem is signifi-
cantly more complicated. Rigorous scattering theory for a system of three
particles valid for short-range potentials was given by Faddeev [17, 18]. For
the charged particles with the long-range Coulomb interaction the theory has
faced difficulties associated with the compactness of the underlying equations.
A renormalization method was implemented successfully for the three-body
problem when only two particles are charged [19]. There are no compact in-
tegral equations yet known for collisions of more than two charged particles
that are satisfactory above the breakup threshold [16]. Furthermore, there
is no theoretical proof or practical evidence that a renormalization approach
can be applied to the Faddeev equations for the three-body Coulomb prob-
lem. This has serious consequences especially for three-body problems in
atomic physics where all three particles are charged.
As far as breakup of a bound state of two particles in a system of three
charged particles is concerned, the key problem is how to extract the scatter-
ing information from the wavefunction when the latter is available. The
conventional theory fails to provide a formal post-form definition of the
breakup amplitude for three charged particles in terms of the total wave-
function with outgoing scattered waves describing the process. Therefore,
the Coulomb interaction is screened and the formula for the short-range case
is used. However, it is well known that the short-range definition of the
breakup amplitude diverges when the screening radius is taken to infinity.

5
Thus we have a situation where we cannot use the theory unless we screen
the Coulomb interaction, and when we do, we end up with quantities which
diverge as screening is removed. This leaves no choice but to use renormal-
ization to fix unphysical results. As mentioned above this is not possible in
all cases of interest. Therefore, a new approach to the Coulomb three-body
problem that does not need renormalization is required.
Despite the aforementioned problems remarkable progress has been achieved
in atomic physics in describing (e,2e) processes via the exterior complex-
scaling (ECS) [1, 2] and the convergent close-coupling (CCC) [4] methods.
As we mentioned earlier, the success of the ECS and CCC approaches to
the Coulomb breakup problem caused us to reexamine the underlying formal
theory [12–15]. In the ECS method the amplitude is calculated from Pe-
terkop’s trial integral [20] which is likely to be some kind of approximation
to the exact breakup amplitude in the post form (unavailable in conventional
scattering theory). In the CCC method one of the electrons is treated using
a square-integrable (L2 ) representation, and the breakup amplitude can be
related to a particular form of Peterkop’s integral. Despite the success of
the computational methods in describing the measured cross sections, the
traditional formal theory of scattering was unable to provide a definition
for the breakup amplitude in terms of the scattered wave. This has been a
long-standing problem [14].
In this Section we briefly review a surface-integral approach [15] to for-
mulating scattering theory. This approach represents an extension of the
general formalism of scattering theory to systems of two and three charged
particles with long-range Coulombic interactions. The essential feature of the
surface-integral formulation of scattering theory is that it avoids the reference
to the Green’s function and formal solution for the scattered wavefunction in
the integral form. Wave functions we deal with in scattering theory go be-
yond the Hilbert space. Since these functions are not square-integrable (L2 )
their scalar products can be unbounded. While this fact is not a problem on
its own, nevertheless, non-L2 functions do make certain integrals emerging
in the theory divergent. In case of integrals containing the interaction po-
tential a standard procedure, which ensures their existence, is limiting the
range of the potential. However, this irreversibly distorts the nature of the
problem. The surface-integral approach usees a different way to deal with
the aforementioned problem. Here, first the scattering problem is formulated
in a finite region of coordinate space, and then extended to the full space.
This leads to new more general definitions for scattering amplitude valid for

6
arbitrary interactions including Coulombic long-range ones. Details of the
surface-integral formalism can be found in Ref. [15].

2.1. Formal problems of the conventional theory of ionisation

We start from giving some more details of two problems of formal scatter-
ing theory mentioned above. Consider scattering of electron e1 with incident
momentum ki off a hydrogen atom (p, e2 ) in initial state φi (r2 ) of energy
Ei . Assume that the energy of the projectile ki2 /2 is enough to break up the
target. The ionization amplitude in the prior form is well defined and given
according to [21]
Z
T (k1 , k2 ) = dr1dr2 Ψ−∗
0 (r1 , r2 )V i Φi (r1 , r2 ), (1)

where Ψ− 0 is the total scattering wave function developing into an initial state
of three particles in the continuum with incoming scattered-wave boundary
condition and describes

e1 + e2 + p 
e1 + (p, e2 ) → e1 + e2 + p (2)

e2 + (p, e1 )

processes. The wave function Ψ−

0 satisfies the Schrödinger equation

(E − H)Ψ−
0 (r1 , r2 ) = 0, (3)

where H = H0 +V is the total three-body Hamiltonian, H0 = −∆r1 /2−∆r2 /2

is the free three-body Hamiltonian, V is the full interaction and E = ki2 /2 +
Ei = k12 /2 + k22 /2 is the total energy of the system, V i = V − Vi in Eq. (1) is
the interaction of the incident electron with the target particles, r1 and r2 are
the coordinates of the electrons relative to the proton and k1 and k2 are their
momenta. The wave function representing the initial two-fragment channel
is given by a product of the incident plane wave and the initial bound-state
wave function

Φi (r1 , r2 ) = eiki ·r1 φi (r2 ). (4)

Vi is the potential responsible for the bound state in the initial channel.
According to our particular choice of the initial channel, Vi is the Coulomb
interaction of electron e2 and the proton.

7
 The ionization amplitude given in the form (1) is difficult to calculate
because it requires the total scattering wave function Ψ− 0 , which evolves into
a free three-particle state (in a time-reversal picture). Asymptotic boundary
condition that helps to choose the right solution from the infinite set of
solutions to the differential equation has a complicated form. For this reason,
it is very difficult to solve Eq. (3) using direct integration methods. In
addition, for the ionization amplitude to be calculated from this definition,
a knowledge of Ψ− 0 in the entire space is necessary. Therefore, this form of
the ionization amplitude has mostly been used in distorted-wave Born-type
calculations (see, e.g. [22–25] and references therein). In terms of numerical
calculations, it is always more convenient to work with the wave function Ψ+ i
that describes

 e1 + e2 + p
e1 + (p, e2 ) → e1 + (p, e2 ) (5)

e2 + (p, e1 )

processes. The total scattering wave function Ψ+

i is a solution of the Schrödinger
equation

(E − H)Ψ+
i (r1 , r2 ) = 0, (6)

with outgoing scattered-wave boundary condition. The advantage of Ψ+ i over

− +
its time-reversal counterpart Ψ0 is that the wave function Ψi develops from
the exact initial state Φi given by a product of a plane wave and hydrogen
bound state wave function. In order to calculate the ionization amplitude
from Ψ+ i we need a post-form definition of the amplitude. According to the
conventional scattering theory this forms is written as
Z
T (k1 , k2 ) = dr1 dr2Φ∗0 (r1 , r2 )V Ψ+
i (r1 , r2 ), (7)

where

Φ0 (r1 , r2 ) = eik1 ·r1 +ik2 ·r2 . (8)

I.e., the final state of the two unbound electrons is described by the plane
waves. The problem with the definition (7) is that it is valid only for parti-
cles interacting via short-range potentials. It cannot be applied for charged
particles with the long-range Coulomb interaction. An attempt to apply it to

8
breakup processes with charged particles leads to divergent results. As men-
tioned in the Introduction, the prescription from the conventional scattering
theory is to screen the Coulomb interactions, use Eq. (7) and then in the
obtained result take the screening radius to infinity. However, this procedure
has not been shown to be valid neither theoretically, nor by practical calcu-
lations. Then the question is how do we calculate the ionisation amplitude
if we somehow managed to solve Eq. (6) and obtained Ψ+ i ?
Trying to resolve this problem Peterkop [26] (see also [27]) developed 50
years ago the so-called effective-charge formalism for electron-impact ionisa-
tion. He considered the integral
Z
Iz1 ,z2 (k1 , k2 ) = dr1 dr2 Ψ+ (2C)−∗
i (r1 , r2 )(H − E)Φz1 ,z2 (r1 , r2 ), (9)

(2C)−
where the function Φz1 ,z2 is a product of the two Coulomb (2C) wave func-
tions with effective charges z1 and z2 :

Φ(2C)−
z1 ,z2 (r1 , r2 ) = e
ik1 ·r1 +ik2 ·r2
ψz1 (k1 , r1 ) ψz2 (k2 , r2 ), (10)

with incoming wave boundary condition. The Coulomb part is given by

ψν (k, r) = Γ(1 + iν/k)exp(πν/2k) 1F1 (−iν/k, 1; −i(kr + k · r)), (11)

where 1F1 is the confluent hypergeometric function.

Using Eq. (6) and Green’s theorem [28] the volume integral in Eq. (9)
can be written as a surface integral
Z Z π/2
1
Iz1 ,z2 (k1 , k2 ) = lim R5 dr̂1dr̂2 dα sin2 α cos2 α
2 R→∞ 0
" #
(2C)−∗
∂Ψ+
i + ∂Φz1 ,z2
× Φ(2C)−∗
z1 ,z2 − Ψi , (12)
∂R ∂R

where R = (r12 + r22 )1/2 is a hyperradius, (r̂1 , r̂2, α) is a 5-dimensional hyper-

angle, with α = arctan(r2 /r1 ).
The advantage of the integral form (12) is that it is readily expanded in
partial-waves leading to a sum of one-dimensional integrals. On the other
hand Eq. (1) reduces to a two-dimensional integral upon partial wave expan-
sion. Most importantly, the integral Iz1 ,z2 depends only on the asymptotic

9
 (2C)−
behavior of the wave functions Ψ+ i and Φz1 ,z2 on an infinitely large hyper-
shpere and, therefore, knowledge of the wave function Ψ+ i over the entire
space is not required. But how does integral Iz1 ,z2 relate to the ionization
amplitude?
Let us define the domain Ω0 to correspond to the space where all inter-
particle distances are large, i.e. r1 , r2 , r3 (r3 = r1 − r2 ) → ∞ , in a manner
that r1 /r2 → const6= 0. In this domain the asymptotic behavior of Ψ+ i was
found by Peterkop [29] and is written, in the leading order of R, as
Ω
Ψ+ 0
i (r1 , r2 ) −→ A(r̂1 , r̂2 , α) R
−5/2 iκR+iγ ln(κR)
e , (13)

where κ = (2E)1/2 ,
 
1 1 1 1
γ= + −√ , (14)
κ cos α sin α 1 − r̂1 · r̂2 sin 2α
and A is Peterkop’s ionization amplitude. Peterkop showed that integral
Iz1 ,z2 exists and differs from amplitude A only by a factor:

κ3/2 iβ(R)+iπ/4
A(k̂1 , k̂2 , α′ ) = e Iz1 ,z2 (k1 , k2 ), (15)
(2π)5/2
where α′ = arctan(k2 /k1 ). However, the phase factor β(R) diverges as R →
∞ unless the so-called Peterkop condition
z1 z2 1 1 1
+ = + − , (16)
k1 k2 k1 k2 |k1 − k2 |
is satisfied. In this case β(R) vanishes for large R. The relation (15) is known
as the Peterkop integral representation for the ionization amplitude.
Thus, in Peterkop’s effective charge approach z1 and z2 depend on vectors
k1 and k2 . For this reason this method was not very useful in practice for
it was not clear how to implement condition (16) in realistic calculations.
Numerical problems associated with the effective charge approach were dis-
cussed by McCurdy et al. [30]. Their calculations showed that use of effective
charges z1 and z2 leads to severe numerical problems due to nonorthogonal-
ity of the Coulomb wave of a non-unit effective charge to the bound states
of hydrogen. From a formal point of view, even if the Peterkop condition
is satisfied, one could not establish the ionization amplitude in full. This is
because, as mentioned by Peterkop [27], an arbitrary part of the complex

10
amplitude A in asymptotic form (13) can be moved to the phase factor and
the resulting wave function would still be a solution to the original Eq. (6)
transformed into the six-dimensional hyperspace. Thus, the remaining part
of A can equally well be called an ionization amplitude and there is no way
of choosing between the different phase possibilities which is clearly unsatis-
factory.
In order to explain the origin of the problems with the Peterkop formula-
tion, summarized by Eqs. (12) and (15), we need to distinguish all possible
geometries where the condition R → ∞ is satisfied. In addition to the Ω0
domain defined earlier, we identify the domain where r1 → ∞, r2 → ∞ with
limited r3 as Ω3 and when r2 (or r1 ) goes to infinity but r1 (r2 ) remains
limited as Ω2 (Ω1 ). Domains Ω1 , Ω2 and Ω3 correspond to α → π/2, α → 0
and α → π/4 in the surface integral (12), respectively. The problems with
Peterkop’s integral representation for the ionization amplitude originate from
the fact that the Peterkop asymptotic form used to calculate the integral (12)
is valid only in Ω0 . It is clearly seen from Eq. (14) that the Peterkop form
cannot be used when α → 0 and α → π/2. This wave function is singular
also when α → π/4 if r̂1 · r̂2 = 1. At the same time integration over α runs
through all these points. Thus, in the integral representation suggested by
Peterkop the contributions from Ω1 , Ω2 and Ω3 domains are either missing,
or taken into account incorrectly.
Since the Peterkop condition turned out to be impossible to satisfy in
practice, for almost four decades the formalism had looked like an elaborate
theory without practical implications. A breakthrough came in the form of
ECS method [30]. These authors practically demonstrated that for calcu-
lations of the cross sections using the integral representation of Eq. (15)
for the breakup amplitude there was no need to satisfy the associated Pe-
terkop condition. Moreover, they showed that in calculations it was more
convenient to choose the trial function Φ2C z1 ,z2 in Eq. (9) as a combination of
two two-particle scattering states. In the hydrogen ionization problem this
corresponds to taking z1 = z2 = 1 in Eq. (10). Since these two-particle
scattering states are orthogonal to bound-state wavefunctions of the rele-
vant pair of particles, this lead to faster convergence of the resulting integral
[30]. Despite this breakthrough in practical calculations the origin of the
Peterkop integral remained unclear. Also remained ununderstood the reason
why one could ignore the Peterkop condition and still get the correct cross
sections. The latter was addressed in Ref. [13] where the Peterkop formal-
ism was generalised to all domains Ωi , i = 0 − 3, of coordinate space and

11
demonstrated to be valid for arbitrary choice of z1 = z2 = 1, thereby relaxing
the Peterkop condition altogether. This analysis lead to a surface-integral
formulation of the scattering theory which showed that the Peterkop integral
was an approximation to a missing post form of the ionisation amplitude.

2.2. Surface-integral approach

Here we give an overview of the surface-integral approach [15] to scat-
tering theory in a general form applicable for an arbitrary system of three
charged particles. We use the Jacobi coordinates where rα is the relative
coordinate, and kα is the relative momentum, between particles β and γ, ρα
is the relative coordinate of the center of mass of the pair (β, γ) and particle
α, with qα being the canonically conjugate relative momentum. Accordingly,
µα is the reduced mass of particles β and γ and Mα is the reduced mass of
the pair (β, γ) and particle α. Here α, β, γ = 1, 2, 3, and α 6= β 6= γ.
We are interested in scattering of particle α with incident momentum qαn
off a bound pair (β, γ) in initial state φαn (rα ) of energy Eαn . Here n denotes
a full set of quantum numbers of the bound state (β, γ) in channel α. Assume
2
that the energy of the projectile qαn /2Mα is enough to break up the target.
Therefore, in addition to direct scattering and rearrangement (β + (γ, α))
channels, there is a breakup one. We call this 2 → 3 process. In order to
find the amplitudes of direct scattering, rearrangement and breakup in this
collision we need the total scattering wavefunction developed from the initial
channel αn and three different asymptotic wavefunctions corresponding to
three final-state channels. The same amplitudes can be found in the so-
called prior forms as well, which requires the knowledge of the other three
types of the total scattering wavefunctions being developed to three different
final-state wavefunctions. Thus, in any case, we require a set of four total
scattering wavefunctions together with their corresponding asymptotic forms
in all relevant asymptotic domains.
There are two distinct types of asymptotic domains. Let us call Ω0 the
asymptotic domain, where all interparticle distances are large, i.e, rα → ∞,
ρα → ∞, so that rα /ρα is non-zero. In addition, we call Ωα the asymptotic
regime, where ρα → ∞, however rα satisfies the constraint rα /ρα → 0.
The total three-body wavefunction describing the 2 → 3 processes satis-
fies the Schrödinger equation

(E − H)Ψ+
αn (rα , ρα ) = 0, (17)

12
with outgoing-wave boundary conditions, where H = H0 +V , H0 = −∆rα /2µα−
∆ρα /2Mα is the free Hamiltonian, V = Vα (rα ) + Vβ (rβ ) + Vγ (rγ ) is the full
2
interaction and E = Eαn + qαn /2Mα = kα2 /2µα + qα2 /2Mα is the total energy
of the system. We split the wavefunction Ψ+ αn into the initial-channel wave
Φ+αn and scattered wave Ψsc+
αn :

Ψ+ + sc+
αn (rα , ρα ) = Φαn (rα , ρα ) + Ψαn (rα , ρα ). (18)

A slightly different scattering process may also take place within the
same three-body system at the same total energy E, the one where in the
initial channel (in the time-reversed picture this will be the final state) all
three particles are in the continuum which we call 3 → 3 scattering. The
wavefunction Ψ− 0 describing this process is also an eigenstate of the same
Hamiltonian H,

(E − H)Ψ−
0 (rα , ρα ) = 0, (19)

but with incoming scattered-wave boundary conditions. In the total wave-

function Ψ−0 (rα , ρα ) we separate the part describing the unscattered state of
three free particles, denoted Φ− −
0 (rα , ρα ) and which Ψ0 is being developed to
(in the absence of the Coulomb interaction this would simply be the three-
body plane wave),

Ψ− − sc−
0 (rα , ρα ) = Φ0 (rα , ρα ) + Ψ0 (rα , ρα ). (20)
−
Asymptotic boundary conditions that the wavefunctions Ψ+ αn and Ψ0 satisfy
are compiled in Ref. [15].
−
Now we use Ψ+ αn and Ψ0 to derive amplitudes for different scattering pro-
cesses. This time we need an incomplete inner product in the six-dimensional
configuration space. Such a product of two arbitrary functions Ψi and Ψf is
written as a volume integral
Z
Ψf |Ψi R0 = drα dρα Ψ∗f (rβ , ρβ )Ψi (rα , ρα ), (21)
R≤R0

where the integration is limited to the volume of a six-dimensional hyper-

sphere of radius R0 . A hyperradius in the six-dimensional configurations
1/2
space is defined according to R = (µα rα2 + Mα ρ2α ) .
Taking into account Eq. (18) we can write Eq. (17) as

(E − H)Ψsc+ +
αn (rα , ρα ) = (H − E)Φαn (rα , ρα ). (22)

13
Let us multiply Eq. (22) by Ψ−∗
0 (rα , ρα ) from the left and integrate the result
over the volume of a hypersphere of radius R0 :

Ψ− sc+
0 |(E − H)Ψαn R0
= Ψ− +
0 |(H − E)Φαn R0
. (23)

We also have

(E − H)Ψ− sc+
0 |Ψαn R0
= 0, (24)

which is true for any R0 simply due to Eq. (19). Now we subtract Eq. (24)
from (23) to get

Ψ− sc+
0 |(E − H)Ψαn R0
− (E − H)Ψ− sc+
0 |Ψαn R0
= Ψ− +
0 |(H − E)Φαn R0
. (25)

Despite of the fact that both Ψ− sc+ 2

0 and Ψαn are non-L functions, terms of the
− sc+
form Ψ0 |(E − V )|Ψαn R0 are finite due to the limited space (regardless of
the long-range nature of the potential). Therefore, canceling them we look
at the limit of this equation as R0 → ∞
 
lim − Ψ− sc+
0 |H0 Ψαn + H 0 Ψ− sc+
0 |Ψαn R0
= lim Ψ− +
0 |(H − E)Φαn R0 .(26)
R0 →∞ R0 →∞

As in the two-body case, the meaning of this quantity will become clear when
we evaluate the limits.
Parameter R0 can go to infinity with the system being in Ω0 or Ωα ,
α = 1, 2, 3. An essential feature of the term on the LHS of Eq. (26) is that
it is easily transformed into an integral over the hypersurface of radius R0
so that the result depends only on the behavior of the wavefunctions on this
surface. For this integral the knowledge of the wavefunctions anywhere inside
the surface is not required. Then it can be evaluated using the asymptotic
forms of the wavefunctions in the corresponding asymptotic domain.
Let us start with the Ω0 domain. If R0 → ∞ in Ω0 then for the LHS of
Eq. (26) we have
Z Z π/2
1 5
LHS = lim R dr̂αdρ̂α dϕα sin2 ϕα cos2 ϕα
2(µαMα )3/2 R0 →∞ 0 0
 
−∗ ∂ sc+ sc+ ∂ −∗
× Ψ0 (rα , ρα ) Ψ (rα , ρα ) − Ψαn (rα , ρα ) Ψ (rα , ρα ) ,
∂R αn ∂R 0 R=R0
(27)

14
 h i
where ϕα = arctan (µα /Mα )1/2 rα /ρα is a hyperangle, 0 ≤ ϕα ≤ π/2. Here
we first transformed H0 from (rα , ρα ) into (R, rbα , ρbα , ϕα )-variables and then
made use of Green’s theorem to transform the volume integral into the surface
integral. Now we can use the asymptotic forms for the wave functions [15]
and perform differentiation to find out that this is an extremely oscillatory
integral as R0 → ∞. Therefore, only points of stationary phase in ϕα , if
there are any, should contribute to the integral. Calculating the integral by
means of the stationary-phase method we arrive at

LHS = T (kα , qα ) , (28)

where T (kα , qα ) is the amplitude of the of the scattered wave in Ω0 . There-

fore, in Ω0 Eq. (26) is written as

T (kα , qα ) = lim Ψ− +
0 |(H − E)Φαn R0
. (29)
R0 →∞

In other words, if scattering takes place into the Ω0 domain then the expres-
sion limR0 →∞ Ψ− +
0 |(H − E)Φαn R0 represents the breakup amplitude. If after
the collision the products of scattering turn out to be in Ωα or in Ωβ domains
then we have to distinguish whether all three particles are in the continuum
or only one. If all three are in the continuum then similarlarly to the Ω0 case
we can show that limR0 →∞ Ψ− +
0 |(H − E)Φαn R0 again represents the breakup
amplitude. Thus, Eq. (29) defines the breakup amplitude in all asymptotic
domains.
If after the collision the products of scattering form a two-fragment chan-
nel then instead of Ψ− 0 we will need the total scattering wavefunction which
develops into the wavefunction of this two-fragment channel. We start from
Ωα domain which corresponds to direct scattering. In this case the total scat-
tering wavefunction we need is Ψ− −∗
αm . Let us multiply Eq. (22) by Ψαm (rα , ρα )
from the left and integrate the result over the volume of a hypersphere of
radius R0 :

Ψ− sc+
αm |(E − H)Ψαn R0
= Ψ− +
αm |(H − E)Φαn R0
. (30)

Now we subtract (E − H)Ψ− sc+

αm |Ψαn R0
= 0 from Eq. (30) to get

− Ψ− sc+
αm |H0 Ψαn R0
+ H 0 Ψ− sc+
αm |Ψαn R0
= Ψ− +
αm |(H − E)Φαn R0
. (31)

15
We again consider the limit of this equation as R0 → ∞. Since this time
R0 → ∞ in Ωα then on the LHS of Eq. (31) we have
Z 
1 ∂ sc+
LHS = lim R0 drα dρ̂α Ψ−∗
2
αm (rα , ρα ) Ψ (rα , ρα )
2Mα R0 →∞ ∂ρα αn

sc+ ∂ −∗
−Ψαn (rα , ρα ) Ψ (rα , ρα ) . (32)
∂ρα αm ρα =R0

Here we transformed only one of the volume integrals into the surface integral
(the wave functions fall off exponentially in the other two-body subspace).
Using the asymptotic forms of the wavefunctions [15] and the orthogonality
of the the two-particle bound state wavefunctions and calculating the integral
we arrive at

LHS = F (qαm , qαn ), (33)

where F (qαm , qαn ) is the amplitude of the wave scattered into channel α.
Thus as R0 → ∞ Eq. (31) is in fact written as

F (qαm , qαn ) = lim Ψ− +

αm |(H − E)Φαn R0
. (34)
R0 →∞

In other words we have got a definition for the direct scattering (elastic and
excitation) amplitude. Finally, taking R0 → ∞ in Ωβ (i.e., the final state
belongs to channel β) and calculating the limit of Eq. (31) we get a definition
for the amplitude of the rearrangement scattering

G(qβm , qαn ) = lim Ψ− +

βm |(H − E)Φαn R0
. (35)
R0 →∞

Now we derive the scattering and breakup amplitudes in post form. Tak-
ing into account Eq. (20) we can write Eq. (19) as

(E − H)Ψsc− −
0 (rα , ρα ) = (H − E)Φ0 (rα , ρα ). (36)

Let us take the complex conjugate of Eq. (36) and multiply it by Ψ+

αn (rα , ρα )
from right. Then integrating the result over the volume of a hypersphere of
radius R0 we get

(E − H)Ψsc− +
0 |Ψαn R0
= (H − E)Φ− +
0 |Ψαn R0
. (37)

16
We also consider
Ψsc− +
0 |(E − H)Ψαn R0
= 0, (38)
which is again valid for any R0 due to Eq. (17). Now we subtract Eq. (38)
from (37)
(E − H)Ψsc− +
0 |Ψαn R0
− Ψsc− +
0 |(E − H)Ψαn R0
= (H − E)Φ− +
0 |Ψαn R0
, (39)
which reduces to
− H0 Ψsc− +
0 |Ψαn R0
+ Ψsc− +
0 |H0 Ψαn R0
= (H − E)Φ− +
0 |Ψαn R0
. (40)
Calculations of the limit of the LHS of Eq. (40) are similar to those of Eq.
(26) leading to Eqs. (29), (34) and (35). Depending on whether the R0 → ∞
limit is taken in domains Ω0 , Ωα or Ωβ we have
T (kα , qα ) = lim (H − E)Φ− +
0 |Ψαn R0
, (41)
R0 →∞

F (qαm , qαn ) = lim (H − E)Φ− +

αm |Ψαn R0
, (42)
R0 →∞

G(qβm , qαn ) = lim (H − E)Φ− +

βm |Ψαn R0
, (43)
R0 →∞

respectively. These are the post forms of the amplitudes.

As it will be shown in the following subsection the prior and post forms
of the breakup amplitude can take surface-integral forms (see Eqs. (61) and
(63)). These surface integrals are similar in form to those calculated earlier.
They can be calculated using the same analysis and shown to yield the same
result, i.e. the breakup amplitude. Therefore Eq. (26) and the limit of
(40) as R0 → ∞ represent the breakup amplitude. The same is true for the
surface-integral representations of the direct scattering and rearrangement
amplitudes. Therefore, Eqs. (29), (34) and (35) can in fact be written as
−
→
T pr (kα , qα ) = Ψ− +
0 | H − E|Φαn , (44)
−
→
F pr (qαm , qαn ) = Ψ− +
αm | H − E|Φαn , (45)
−
→
Gpr (qβm , qαn ) = Ψ− +
βm | H − E|Φαn (46)
and Eqs. (41)-(43) as
←−
T pt (kα , qα ) = Φ− +
0 | H − E|Ψαn , (47)
←
−
F pt (qαm , qαn ) = Φ− +
αm | H − E|Ψαn , (48)
←−
Gpt (qβm , qαn ) = Φ− +
βm | H − E|Ψαn . (49)

17
Thus we get alternative representations for the breakup, scattering and rear-
rangement amplitudes. In particular, the definition given in Eq. (47) resolves
the problem of the post form of the Coulomb breakup amplitude mentioned
earlier.
We note that when the interactions between all three pairs are short
ranged then

Φ±
αn (rα , ρα ) → e
iqαn ·ρα
φαn (rα ). (50)

This state satisfies the equation

(H − E)eiqαn ·ρα φαn (rα ) = V α eiqαn ·ρα φαn (rα ), (51)

where V α = V − Vα . At the same time if we have 3 particles in the final

channel then

Φ−
0 (rα , ρα ) → e
ikα ·rα +iqα ·ρα
, (52)

for which we have

(H − E)eikα ·rα+iqα ·ρα = V eikα ·rα+iqα ·ρα . (53)

Therefore, the new generalized forms of the amplitudes (44)-(49) reduce to

the standard definitions

T pr (kα , qα ) = Ψ−
0 |V α |φαn , qαn , (54)
F pr (qαm , qαn ) = Ψ−
αm |V α |φαn , qαn , (55)
pr
G (qβm , qαn ) = Ψ−
βm |V α |φαn , qαn (56)

and

T pt (kα , qα ) = qα , kα |V |Ψ+
αn , (57)
pt
F (qαm , qαn ) = qαm , φαm |V α |Ψ+
αn , (58)
Gpt (qβm , qαn ) = +
qβm , φβm |V β |Ψαn . (59)

When the interactions have the Coulomb tail Eqs. (51) and (53) are not
satisfied. For this reason the standard definitions (54)-(59) become invalid.
As mentioned earlier both prior and post forms of the scattering and
breakup amplitudes take surface integral forms. The post form of the breakup

18
amplitude T pt (kα , qα ) is given by Eq. (41). Using Eq. (17) which is valid
for any R0 we can write this as
 
T pt (kα , qα ) = lim H0 Φ− + − +
0 |Ψαn − Φ0 |H0 Ψαn R0 . (60)
R0 →∞

Transforming the volume integral into a surface integral we get

Z Z π/2
pt 1 5
T (kα , qα ) = − 3/2
lim R0 dr̂α dρ̂α dϕα sin2 ϕα cos2 ϕα
2(µ M ) R →∞
 α α
0 0

+ ∂ −∗ −∗ ∂ +
× Ψαn (rα , ρα ) Φ (rα , ρα ) − Φ0 (rα , ρα ) Ψ (rα , ρα ) .
∂R 0 ∂R αn R=R0
(61)

Thus the breakup amplitude in the post form is written as a five-dimensional

surface integral. In the prior form of this amplitude and also both in the post
and prior forms of scattering and rearrangement amplitudes only one of the
three-dimensional volume integrals can be transformed into a surface integral.
Consider, for instance, the prior form of the breakup amplitude. Using Eq.
(19) we get from Eq. (29)
 − 
T pr (kα , qα ) = lim Ψ0 |H0 Φ+ − +
αn − H0 Ψ0 |Φαn R0 . (62)
R0 →∞

This can be written as

Z
pr 1 2
T (kα , qα ) = − lim R drα dρ̂α
2M R0 →∞ 0
 α
∂ +
× Ψ−∗
0 (rα , ρα ) Φ (rα , ρα )
∂ρα αn

+ ∂ −∗
− Φαn (rα , ρα ) Ψ (rα , ρα ) . (63)
∂ρα 0 ρα =R0

Note that above we took into account the fact that the surface integral in
the other two-body space is zero as rα → ∞ (the function Φ+ αn falls off

19
exponentially in this variable). Similarly we get
Z
pr 1 2
F (qαm , qαn ) = − lim R drα dρ̂α
2Mα R0 →∞ 0

∂ +
× Ψ−∗ αm (rα , ρα ) Φ (rα , ρα )
∂ρα αn

+ ∂ −∗
− Φαn (rα , ρα ) Ψ (rα , ρα ) , (64)
∂ρα αm ρα =R0
Z
pr 1 2
G (qβm , qαn ) = − lim R drα dρ̂α
2Mα R0 →∞ 0

∂ +
× Ψ−∗ βm (rα , ρα ) Φ (rα , ρα )
∂ρα αn

+ ∂ −∗
− Φαn (rα , ρα ) Ψ (rα , ρα ) , (65)
∂ρα βm ρα =R0

and in post forms

Z
pt 1 2
F (qαm , qαn ) = − lim R drα dρ̂α
2Mα R0 →∞ 0

∂ −∗
× Ψ+ αn (rα , ρα ) Φ (rα , ρα )
∂ρα αm

−∗ ∂ +
− Φαm (rα , ρα ) Ψ (rα , ρα ) , (66)
∂ρα αn ρα =R0
Z
pt 1 2
G (qβm , qαn ) = − lim R drβ dρ̂β
2Mβ R0 →∞ 0

∂ −∗
× Ψ+ αn (rβ , ρβ ) Φ (rβ , ρβ )
∂ρβ βm

−∗ ∂ +
− Φβm (rβ , ρβ ) Ψ (rβ , ρβ ) . (67)
∂ρβ αn ρα =R0

What is the advantage of the surface-integral representations over the

volume-integral forms? As we can see from the surface-integral forms for
the amplitudes, they are ideal for practical calculations in partial waves as
integrals over solid angles are taken [13]. Moreover, the corresponding results
depend only on the behavior of the wavefunctions on the hypersurface of ra-
dius R0 , meaning that for these integrals the knowledge of the wavefunctions

20
anywhere inside the surface is not required. Therefore, they can be evalu-
ated using the asymptotic forms of the wavefunctions in the corresponding
asymptotic domain. The main advantage of the surface forms is that the
asymptotic channel wavefunctions Φ− ± −
0 , Φαm and Φβm necessary for extract-
ing the breakup, scattering and rearrangement amplitudes do not have to be
exact. They can be replaced by other suitable functions making sure that
magnitudes of the amplitudes are still calculated exactly. The question is
what properties trial functions, capable of replacing aforementioned three
functions, should posses?
Let us assume that as a result of solving the Schrödinger equation scatter-
ing wave Ψ+ sc+
αn (or Ψαn ) became available. First consider the way of extracting
the direct scattering amplitude. According to Eq. (66), in order to extract
this amplitude we need Φ− αm , or its partial waves. Consider the following trial
surface integral instead
←−
FR0 (qαm , qαn ) = Φαm | H − E|Ψ+
αn R0
, (68)

where

Φαm = eiqαm ρα φαm (rα ). (69)

After some algebra similar to what we have used in the previous section we
arrive at

lim FR0 (qαm , qαn ) = F (qαm , qαn ) lim exp [−iηαm /qαm ln(2qαm R0 )] ,
R0 →∞ R0 →∞
(70)

with a divergent phase. From this result we conclude that

| lim FR0 (qαm , qαn )| = |F (qαm , qαn )|. (71)

R0 →∞

Similarly, for the magnitude of the rearrangement amplitude we get

|G(qαm , qαn )| = | lim GR0 (qαm , qαn )|, (72)

R0 →∞

where
←−
GR0 (qβm , qαn ) = Φαm | H − E|Ψ+
αn R0
, (73)

21
and

Φβm = eiqβm ·ρβ φβm (rβ ). (74)

Finally we consider breakup. The breakup amplitude in terms of Ψ+ αn is

given by Eq. (47) as a volume integral and Eq. (61) as a surface integral.
These forms require Φ−0 . Though the latter is known, its partial waves have
complicated analytical form [13]. Therefore we consider the following surface
integral
←−
IR0 (kα , qα ) = Φ̃0 | H − E|Ψ+
αn R0
, (75)

where Φ̃0 is a trial function. The trial function can be the three-body plane
wave or any other function containing the three-body plane wave as a lead-
ing term at large distances. Another requirement is that it must be easily
expandable in partial waves. Consider the case when

Φ̃0 (rα , ρα ) = eikα ·rα+iqα ·ρα . (76)

Then we get, after some algebra [12, 13],

lim IR0 (qα , qα ) = T (qα , qα ) lim exp [−iλ0 ln(2κR0 ) − iσ0 ] . (77)
R0 →∞ R0 →∞

The phase factor on the RHS diverges logarithmically as R0 → ∞. However,

we can again write that

| lim IR0 (qα , qα )| = |T (qα , qα )|. (78)

R0 →∞

Thus in order to extract the magnitude of the scattering amplitudes it is

not necessary to use the exact asymptotic state. The same can be done
using much simpler trial functions. And that is the main advantage of the
surface-integral representations.
A similar approach to atomic ionization problem is known as the Peterkop
effective-charge formalism [26, 31]. The Peterkop integral discussed earlier
can be written as
←−
Iz1 ,z2 (qα , qα ) = Φ(2C)− +
z1 ,z2 | H − E|Ψαn , (79)
(2C)−
where Φz1 ,z2 is a trial function taken as a product of two Coulomb waves of
effective charges z1 and z2 . Though Eq. (79) has no mathematical meaning

22
as it is written since the RHS is divergent, however, if we assume that it is
written in the sense of Eq. (75) for a finite R0 , then as R0 → ∞ Eq. (78)
would hold.
The CCC [5] and RMPS [7–10] methods are the other two successful ap-
proaches to atomic breakup problem. After inserting the projection operator
into the exact post form of the ionization amplitude these approaches end
up with a representation for the ionisation amplitude similar to the Peterkop
trial integral (see Eqs. (106)-(107) and (118)). The choice of effective poten-
tials in this case corresponds to z1 = 0 for more energetic of the electrons and
z2 = 1 for the other (before the antisymmetrization). Though they do not
use the surface-integral technique for calculating the amplitude, the obtained
representation is further built into resulting scattering equations.
Thus the new post form of the breakup amplitude given in Eq. (47) in
particular explains the origin of the Peterkop integral, a cornerstone of the
highly successful ECS, CCC and R-matrix approaches to Coulomb breakup
problems in atomic physics. Comparison of Eq. (79) with Eq. (47) shows
that the Peterkop integral is simply an approximation to the exact breakup
amplitude in its post form, where the exact three-body state Φ− 0 is replaced by
the trial function Φz1 ,z2 . It is remarkable that with any choice of the effective
charges the difference between the breakup amplitude and Peterkop’s integral
reduces to a phase factor which does not affect the calculated cross sections
[12, 13] provided Ψ+ αn is accurate and R0 is asymptotically large.
In order to solve a scattering problem, first, one has to find the total
wavefunction describing the scattering process. The second part consists in
the extraction of the necessary scattering amplitudes from this wave function
for the purpose of calculating the cross sections. So far the surface-integral
approach has resolved problems of the theory related to extracting the infor-
mation about the scattering event. When the total scattering wavefunction
is available the scattering amplitudes can be reliably extracted from it us-
ing the new definitions regardless the long-range nature of the interactions.
Once the amplitudes are available calculations of corresponding cross sections
are straightforward (see, e.g., [32]). As mentioned earlier, there are sophisti-
cated numerical methods which can provide reliable numerical solution to the
Schrödinger equation in some special cases. However, in case of three distin-
guishable particles where rearrangement is possible the Schrödinger equation
cannot provide a unique answer. This is because of the fact that it is im-
possible to specify all asymptotic boundary conditions using one set of the
Jacobi variables. To overcome this difficulty Faddeev [17] suggested a set

23
of equations which incorporates all the required asymptotic boundary condi-
tions. However, as mentioned earlier in case of charged particles the Faddeev
equations become noncompact. In other words they cannot be solved using
standard numerical techniques though noncompactness generally does not
exclude existence of analytic solutions. At the same time our results show
that problems with the Faddeev equations are more serious than noncom-
pactness. Equivalent sets of the Faddeev equations can be written for the
wavefunction, resolvent of the Green’s function or T-matrix. Let us consider
the equations for the components of the three-body T-matrix. The start-
ing point for these equations are old (conventional) definitions in terms of
the interaction potentials which are simply not correct for charged particles.
This implies that any results derived from old definitions are valid strictly
for short-range potentials. For the Coulombic potentials they might be sim-
ply incorrect. Thus, the Faddeev equations in the presence of long-range
Coulomb interactions require careful inspection. It may be possible to for-
mulate the Faddeev equations in a form that would not require screening and
renormalization along the lines of the present surface-integral formalism.
The main ingredient of the Faddeev equations is the off-the-energy-shell
two-body T-matrix. For charged particles it is the off-shell Coulomb two-
body T-matrix. The theory of the Coulomb T-matrix, which started from
pioneering work of Schwinger in the 1940’s (first published in Ref. [33]), is
believed to be well developed and complete [16]. However, as we mentioned
earlier the off-shell Coulomb T-matrix has no on-shell limit. It diverges as
the on-shell point is approached where physical observables are extracted.
Therefore, it cannot be directly used to calculate the physical Coulomb scat-
tering amplitude. In particular, it appears obvious that if used it can only
exacerbate the problems of the Faddeev equations. The theory is based on
the conception that the Coulomb T-matrix cannot be defined directly on
the energy shell. However, in the present work we have shown this to be a
misconception and introduced the on-shell Coulomb T-matrix which directly
gives the physical scattering amplitude. In other words, the on-shell Coulomb
T-matrix presented here has no singularities and therefore does not require
renormalisation. This means that the conventional theory of the Coulomb
T-matrix must be abandoned altogether and a new theory be developed. We
believe a new off-shell Coulomb T-matrix should be introduced as an analytic
extension into the complex-energy plane of the on-shell T-matrix presented
in this work.
We started with a general comment about divergence problems in different

24
branches of physics, associated with fields creating 1/r potentials. These an-
noying difficulties are consequences of imperfections in the existing theories.
Renormalization is used to deal with such problems in quantum electrody-
namics (QED). QED has experienced extraordinary success since Feynman,
Schwinger and Tomonaga suggested the renormalization method. However,
certain dissatisfaction with this theory remains [34–36]. Berestetskii, Lifshitz
and Pitaevskii [37] start their well-known textbook emphasizing that ”there
is as yet no logically consistent and complete relativistic quantum theory.
... The lack of complete logical consistency in this theory is shown by the
occurrence of divergent expressions when the mathematical formalism is di-
rectly applied, although there are quite well-defined ways of eliminating these
divergences. Nevertheless, such methods remain, to a considerable extent,
semiempirical rules, and our confidence in the correctness of the results is
ultimately based only on their excellent agreement with experiment, not on
the internal consistency or logical ordering of the fundamental principles of
the theory.” While another popular author Ryder [38] concludes that “de-
spite the comparative success of renormalisation theory the feeling remains
that there ought to be a more satisfactory way of doing things.” Unfortu-
nately, these concerns are largely ignored by the scientific community. We
believe there must be a logically consistent solution to these problems where
the mathematical formalism is directly applicable without the need for renor-
malization. In our case, utilising a surface-integral approach and focusing on
integrated properties removes screening and renormalisation requirements in
scattering theory. We suggest that a similar integrated approach may help
eliminate renormalisation requirements in QED and other fields. This looks
feasible at least in the case of the infrared divergencies.

3. Computational methods for Coulomb breakup problems

3.1. Exterior complex scaling
As we have already seen, specifying the asymptotic form of the scattering
wave function for electron–impact ionization is a major challenge due to the
long–range Coulomb force. In order to carry out numerical calculations for
ionization one needs the partial–wave forms as well and they are not in easily
accessible analytic forms amenable to practical calculations. Indeed very few
numerical studies have been carried out, mostly in the Temkin–Poet model
[39–41] that has simpler boundary conditions for the scattered wave. There-
fore alternative methods applying the direct ionization boundary condition

25
have been sought. The CCC approach that has been successfully applied
to ionization [42, 43] has the virtue of replacing continuum waves in the
ionization channel by positive energy pseudo–states that at large distances
approach zero magnitude. In this sense the continuum boundary condition of
three free particles is replaced by one with a single continuum particle and an
excited bound–state with positive energy. Exterior Complex Scaling (ECS)
is another method that also seeks to eliminate the complexity of dealing with
the true ionization boundary condition. In this method whose application to
atomic collisions was championed by Rescigno, McCurdy and co-workers in
the late 1990’s [1, 44, 45] the three-body Schrödinger equation in coordinate
space is solved by rotating the configuration space coordinates into the com-
plex plane at sufficiently large distances r ≥ R0 , where the asymptotic form
of the boundary condition can be employed. For all intents and purposes
R0 may be considered to play the same role as that defined in the previous
section on the formal scattering theory.
The method can be illustrated simply by means of the example of a one–
dimensional scattering problem with a short-range potential U(r). Consider
the radial (s-wave) Schrödinger equation
 2 
d
2
+ k ψk+ (r) = U(r)ψk+ (r).
2
(80)
dr
The scattering solution needs two boundary conditions; as usual ψk+ (0) =
0 and secondly the asymptotic behaviour for large r,

ψk+ (r) ∼ sin kr + eikr eiδ sin δ, (81)

r→∞

where the scattering information is contained in the the phase-shift δ(k). Let
us now split, as in the previous section, ψk+ (r) as the sum of the initial state
φk (r) = sin(kr), and ψksc+ (r) which asymptotically contains the outgoing
scattered wave and has the form

ψksc+ ∼ eikr eiδ sin δ. (82)

r→∞

Now it is possible to express Eq. (80) using this decomposition as

 2 
d
2
+ k − U(r) ψksc+ (r) = U(r)φk (r),
2
(83)
dr

26
In practice this equation for unknown ψ sc+ is easily solved numerically for
short-range potentials using difference equation techniques and propagating
out from r = 0 to large r where the asymptotic form (82) is valid given the
known initial state.
In situations where the asymptotic forms are very difficult to formulate or
to adapt to practical numerical calculations the propagation method floun-
ders. One way to overcome this problem is to use exterior complex scaling.
At a large finite distance from the origin (R0 ) that is in the asymptotic region
we choose to rotate the r–space into the complex plane:
(
r, r < R0
z(r) = iθ (84)
R0 + (r − R0 )e , r ≥ R0 .
The angle of rotation is within the range 0 < θ < π/2 so that when this
transformation is applied to the outgoing wave for r > R0 one has

eikr → eikz(r) = e−k(r−R0 ) sin θ eik(R0 +(r−R0 ) cos θ) (85)

→ 0 as r → ∞. (86)

Thus by using this rotation one can replace the asymptotic scattering
boundary condition by the much simpler one that at some distance not
far beyond R0 , ψksc+ ≈ 0. However there remains a subtlety to consider
when applying this transformation to Eq.(83), namely that the initial state
φk = sin kr has both incoming and outgoing waves. Unfortunately the in-
coming wave diverges exponentially under the transformation. In numerical
implementation it has been found that the empirical procedure of applying
a smooth cut–off to the incoming portion beyond R0 is completely satisfac-
tory if one wishes to extract scattering information in the asymptotic region
below R0 . Mathematical justifications of this empirical procedure has been
provided quite recently [46, 47].
In this example the scattering information, is contained in the coef-
ficient eiδ sin δ of the scattered wave. It can be extracted directly using
eiδ sin δ =r→∞
lim
ψ sc+ e−ikr but one can use the more general (and elegant)
surface integral approach to scattering theory employed in section 2.
Consider the integral
Z  2 
1 R + d 2
tR (k) = ψ (r) + k − U(r) φk (r). (87)
k 0 k dr 2

27
We begin by noting that apart from a kinematic factor it is the partial wave
form of the on–shell post version of the scattering amplitude defined by Eq.
(41) in a one-dimensional case. Making use of Eq.(80) it is straightforward
to derive
1
tR (k) = [ψk+ (R)φ′k (R) − φk (R)ψk+ ′ (R)]. (88)
k
This is essentially the analogue of the surface integral form (61) for a vol-
ume integral of dimensionality one. It is then straightforward to use the
asymptotic forms to deduce that

t(k) = lim tR (k) = eiδ sin δ. (89)

R→∞

There has been considerable testing of the method for ionization in the
Temkin–Poet model [39–41] where there are several approaches that have
provided converged results [48–53]. The ECS approach has demonstrated
it can compute with high accuracy across the full range of electron impact
energies. The method has been successfully applied to the full problem for
ionization [1] as well as discrete scattering [54]. A useful variant of the ap-
proach called the propagating ECS method (PECS) [3, 55] has been applied
to the near threshold ionization region for the full problem and essential el-
ements of the interesting threshold effects predicted by Wannier [56] have
been confirmed [3, 54, 57].
Because the method is so computationally intensive there has been rela-
tively little work carried out beyond the three–body problem. The method
has been applied to four–body problems only in the context of an extended
Temkin–Poet model [58–61].

3.2. Time-dependent close-coupling

Similarly to ECS method the time-dependent close-coupling (TDCS) is a
direct approach to the solution of the three-body Schrödinger equation which
allows to circumvent the difficulties associated with formulating the correct
asymptotic boundary conditions. This, however, is conducted in the time
domain [11, 62–64],

∂Ψ(x1 , x2 , t)
i = H(r1, r2 )Ψ(x1 , x2 , t), (90)
∂t
where the symbol x = (r, s) stands for the combined space r and spin s
coordinates. The total Hamiltonian for electron scattering from hydrogen is

28
given by
1 1 1
H = K1 + K2 − − +− . (91)
r1 r2 |r1 − r2 |
The total wave function is expand in coupled spherical harmonics
X 1 LS
Ψ(x1 , x2 , t) = P (r1 , r2 , t)YlL1 l2 (r̂1 , r̂2 )χS1 1 (92)
Sl1 ,l2
r1 r2 l1 l2 22

where L and S are the total orbital and spin angular momentum. The
coupled spherical harmonic is given by
X
l1 l2 L
Yl1Ll2 (r̂1 , r̂2 ) = Cm Y
1 m2 0 l1 m1
(r̂1 )Yl2 m2 (r̂2 ), (93)
m1 m2

l1 l2 L
where Cm 1 m2 0
is a Clebsh-Gordan coefficient, and Ylm (r̂) is a spherical har-
monic. The coupled two-electron spin wave function χS1 1 is defined similarly
22
to (93). Substitution of the expansion (92) into time-dependent Schrödinger
equation (90) and projection onto coupled spherical harmonics and spin wave
functions leads to a set of coupled partial differential equations for each target
symmetry

∂PlLS (r1 , r2 , t) X
1 l2
i = Tl1 l2 (r1 , r2 )PlLS
l (r 1 , r 2 , t) + UlL1 l2 ,l1′ l2′ PlLS
′ ′ (r1 , r2 , t), (94)
1 l2
∂t 1 2
′ ′
l1 l2

where Tl1 l2 (r1 , r2 ) contains all one-electron operators and UlL1 l2 ,l′ l′ (r1 , r2 ) con-
1 2
tains the electron-electron interaction.
The set of coupled equations (94) can be written as

∂P LS
i = H L P LS , (95)
∂t
where the dimension of vector P LS is equal to the number of l1 l2 pairs. Each
vector component PiLS is represented on a two-dimensional lattice as an array
of dimension N̄ . Similarly, the matrix H L has dimensions of N × N, and
each of its components is an array of dimension N̄ × N̄ . As matrix H L is
time independent the time evolution of the radial wave functions is given
simply by
P LS (t + ∆t) = exp(−iH L ∆t)P LS (t). (96)

29
Choosing the time step ∆t sufficient;y small the radial wave function can be
propagated in time [62, 63]. This allows to obtain a lattice representation of
the radial wave function.
At time t = 0 the wave function is constructed as
1

PlLS
1 l2
(r1 , r2 , t = 0) = √ Gk1 l1 (r1 )P1s (r2 ) + (−1)S δ0,l1 Gk2 l2 (r2 )P1s (r1 ),
2
(97)
where Gkl (r) is a radial wave-packet for linear momentum k [63]. At an
appropriate time t = T after the collision, when only an outgoing wave
is present for each reaction channel, the ionization scattering amplitude is
simply defined as the overlap between two Coulomb waves and the calculated
two-electron wave function,
(−) (−)
f (S) (q1 , q2 ) = hq1 q2 χS1 1 |Ψ(t = T )i. (98)
22

We cannot offer a formal proof as to why Eq. (98) should yield the cor-
rect ionization amplitude. It is clear that the projection by the two non-
interacting Coulomb waves eliminates any orthogonal components in the to-
tal wavefunction. We suppose that Eq. (98) is a good approximation to the
standard interpretation in terms of a probability amplitude of finding the sys-
tem in a state with the electrons having the desired momenta. The TDCS
method was successfully applied to the calculation of fully-differential cross
sections for electron impact ionization of hydrogen [11, 64] and helium [65, 66]
atoms.

3.3. Convergent close-coupling

The convergent close-coupling (CCC) method is a technique for treating
a projectile-target collision problem. The aim is to solve such systems at any
collision energy for the major scattering and ionization processes. Initially,
the method was implemented for the simplest well-studied Coulomb three-
body problem of electron scattering on atomic hydrogen for excitation [42]
and total ionization [43]. A few years later it was shown to work for fully
differential ionization as well [5]. The technique has been generalised to other
projectiles, including photons [67], positrons [68, 69], and more recently to
heavy projectiles such as antiprotons [70].
The fundamental strength of the method is the rigorous treatment of
the atomic target countably infinite discrete spectrum and its uncountably

30
infinite continuum in the evaluation of collision data. Specifically, the target
spectrum is expanded using a complete Laguerre basis
 1/2
λl (k − 1)!
ξkl (r) = (λl r)l+1 exp(−λl r/2)Lk−1
2l+2
(λl r), (99)
(2l + 1 + k)!

where the L2l+2

k−1 (λl r) are the associated Laguerre polynomials, and k ranges
from 1 to the basis size Nl . This ensures that with increasing basis size
the negative-energy states converge to the true eigenstates |φn i, while the
positive-energy states provide for an increasingly dense discretisation of the
continuum. Most importantly, the summation over all of the square-integrable
Laguerre-based target states provides a quadrature rule for the summation
and integration over the true target discrete and continuous eigenstates [71].
Specifically, we write
Z
X N
X
I= |φn ihφn | = lim |φ(N ) (N )
n ihφn | ≡ lim I
(N )
. (100)
N →∞ N →∞
n n=1

The target states are typically obtained by diagonalising the target Hamil-
tonian HT to yeild
(N ) (N ) (N )
hφf |HT |φi i = ǫf δf i . (101)
Another alternative is to generate Box-based eigenstates by generating a
(R )
discrete set of states φn 0 (r) which are zero at a specified value of r = R0 ,
and then set to be zero for r > R0 . Being eigenstates they also satisfy
Eq. (101), and so may readily be used in the CCC formalism instead of the
Laguerre-based states [72]. Increasing R0 allows for more discrete eigenstates
to fit in the box and increases the density of positive-energy states. Whereas
in the Laguerre basis the maximum number of states for each l is Nl , in the
box-basis case there is no relation between Nl and R0 . For any specified R0
we can take as many Nl as we like.
(+)
The target states are used to expand the total wavefunction |Ψi i of the
projectile-atom scattering system, i.e.
(+) (+)
0 = (E − H)|Ψi i ≈ (E − H)I (N ) |Ψi i
XN
) (N +)
= (E − H) |φ(N
n fni i
n=1
(N +)
≡ (E − H)|Ψi i, (102)

31
where E and H are the total energy and Hamiltonian, respectively. The
(N +)
unknown one-electron functions fni (r) are to be determined from solution
of the resulting close-coupling equations. The key idea here is that with
increasing N the expansion approaches the identity operator I.
The close-coupling equations can be written in several forms, as discussed
(N )
above, all of which should yield identical results if the same states |φn i are
used in the expansion. In the CCC method the equations take the form of
coupled Lippmann-Schwinger equations for the transition amplitude
(N ) (N ) (N ) (N )
hkf φf |T |φi ki i = hkf φf |V |φi ki i
XN Z (N ) (N ) (N ) (N )
3
hkf φf |V |φn kihkφn |T |φi ki i
+ dk (N )
(103)
,
n=1 E + i0 − ǫn − k 2 /2
where E = ǫi +ki2 for projectile of incident momentum ki on the initial target
state of energy ǫi . The operator V contains all particle interactions, as well
as ensuring the required symmetry properties for the total wavefunction [42].
Without loss of generality in the following we proceed by ignoring sym-
metrisation parts of V and write
H = Hasym + V
= HT + K0 + V, (104)
where the asymptotic Hamitonian is a combination of the target Hamiltonian
and the projectile kinetic energy operator K0 . The transition amplitudes are
then
(N ) (N ) (N ) ←
− (N +)
hkf φf |T |φi ki i = hkf φf | H − E|Ψi i
(N ) (N +)
= hkf φf |V |Ψi i (105)
(N )
for ǫf < E, i.e. all open states. For negative-energy final eigenstates these
are the required scattering amplitudes. In the case of ionization we start
with Eq.(47)
←− (+) (N ) ←− (+)
hΦ−
0 | H − E|Ψi i ≈ hΦf |I ( H − E)I (N ) |Ψi i
(−) ←− (N +)
= hkf qf |I (N ) ( H − E)|Ψi i
N
X (−)
(N ) ← − (N +)
= hqf |φ(N )
n ihkf φn |( H − E)|Ψi i (106)
n=1
(−) (N ) (N ) (N +) (N )
= hqf |φf ihkf φf |V |Ψi i, for qf2 /2 = ǫn ,

32
 (−) (N ) (N )
= hqf |φf ihkf φf |T |φiki i
(N )
≡ fi (qf , kf ), (107)
(−)
where we set hΦf | = hkf qf | because in the close-coupling approximation
asymptotically the projectile electron hkf | is always a plane wave due to being
shielded by the square-integrable target states, ensuring the other electron
(−)
being a pure Coulomb wave hqf |. The summation in (106) only disappears
(N ) (−) (N ) (−) (N )
whenever qf2 /2 = ǫn for some n, as then hqf |φn i = δf n hqf |φf i. In
practice, we evaluate (107) at all values of f and interpolate onto the experi-
mental values. Note that Bray and Fursa [73] derived (107) without knowing
the origin of (47), and that there is a pure Coulomb phase associated only
with the electron of momentum qf .
Thus we have a very simple way of generating the ionization amplitudes
from the close-coupling formalism. We just take the amplitudes for excitation
of the positive-energy states, and multiply them by the given overlap, which
may be seen as restoring the continuum normalisation. The first attempt at
applying the close-coupling formalism to ionization, by Curran and Walters
[74], which was followed by Bray et al. [75], had the perspective of using the
close-coupling expansion to reconstruct the total wavefunction. This runs
into problems when the channel functions are not consistent with the close-
coupling expansion leading to non-existent matrix elements. For this reason
we abandoned such an approach and developed the one above, which turned
out to be much simpler and internally consistent.

3.4. R-matrix with pseudo-states

Application of the R-matrix method to electron scattering from atoms
was originally aimed to study low energy processes [76, 77]. The approach
starts with the multichannel expansion of the total wave function
N
X X
Ψ(x1 , ..., xM +1 ) = A φi (x1 , ..., xM )Fi + cj ξj (x1 , ..., xM +1 ). (108)
i j

Here M is the number of electrons in the target atom, φi (x1 , ..., xM ) are
the target wave functions, Fi are corresponding channel functions, and N is
number of the target states included in the close-coupling expansion. The
ξ(x1 , ..., xM +1 ) are short-ranged functions describing the important correla-
tions that may be missing in the first term of (108).

33
 Substitution of the expansion (108) into the Schrödinger equation, and
projecting on target states φi and correlation functions ξj , leads to formu-
lation of a set of close-coupling integro-differential equations for the channel
functions
 2  X
d li (li + 1) 2Z 2
− + + k i Fi (r) = (Vij + Wij )Fj (r), (109)
dr 2 r2 r j

where Vij represents a local direct potential and Wij represents a non-local
exchange and correlation potential.
The R-matrix method offers an efficient way of solution of the set of
close-coupling equations (109). It makes use of a separate treatment of two
regions; in the inner region (0 < r < a) interaction between target and
projectile electrons is accounted for without approximations while in the
outer region (r > a) only the direct interaction is retained. Solution of Eq.
(109) in the outer region can be obtained relatively easily [78]. Matching
solution in the inner and outer region at the boundary r = a allows to obtain
the scattering information. For na open channels, functions Fij have the
following asymptotic form
−1/2
Fij (r) ∼ ki (Fi + Gi Kij ), i, j = 1, ..., na .
r→∞
(110)
Here Fi and Gi are regular and irregular Riccarti-Bessel functions, respec-
tively, and Kij is the reaction K-matrix.
In the inner region the solution of the Schrödinger equation for the total
wave function at energy E is sought as an expansion in an energy-independent
basis set X
ΨE = AEk Ψk . (111)
k
The basis set Ψk is constructed as
X X
Ψk (x1 , ..., xM +1 ) = A φi (x1 , ..., xM )uj (xM +1 )aijk + ξj (x1 , ..., xM +1 )bjk ,
ij j
(112)
where uj are continuum functions describing the projectile electron. A dis-
crete set of such functions at energies ej is obtained as solutions of single-
particle Schrödinger equation that satisfy the logarithmic derivative bound-
ary condition at the R-matrix boundary
a duj
= b, (113)
uj dr r=a

34
where b is an arbitrary constant. Coefficients aijk , bjk as well as associated
energies Ek are determined by diagonalizing the total projectile and target
electron Hamiltonian
hΨk |H|Ψni = Ek δkn . (114)
Following the standard R-matrix formulation, we define
X
Fik (r) = uj (r)aijk (115)
j

Then the radial function describing the projectile electron in channel i at

energy E is X
Fi = AEk Fik . (116)
k

One can show that coefficients AEk can be expressed as [76]

X  
1 dFi
AEk = Fik a − bFi . (117)
2a(Ek − E) i dr |r=a

In order to obtain the reaction K-matrix the matching at the R-matrix

boundary r = a is performed by substituting the asymptotic form (110)
into left and right-hand sides of Eq. (116). If the R-matrix radius a is not
in the asymptotic region for the potential Vij the set of linear independent
fundamental solutions of Eq. (109) have to be used instead of Riccarti-Bessel
functions Fi and Gi . This leads to a system of linear equations for reaction
K-matrix, and when solved the K, S, and T-matrix for excitation of various
target states become available.
Computationally, the most demanding step in the R-matrix method is the
setting up of the Hamiltonian matrix (114) and its digonalization. The size
of the Hamiltonian matrix grows rapidly with the increase of the R-matrix
radius a, maximum collision energy of interest, and number of target states
included in the close-coupling expansion. Traditionally the R-matrix method
was applied at collision energies below the ionization threshold. However,
with the substantial growth of computational resources over last two decades
the target state expansion in Eq. (112) can be expanded to include a large
number of target pseudo-states [7–9]. Such a modification of the R-matrix
method allowed its application at energies above the ionization threshold,
and it became known as the R-matrix with pseudo-states method. As in
the CCC method, it makes use of a square-integrable representation of the

35
target atom continuum, and yields the scattering T-matrix for excitation of
the positive-energy pseudo-states.
(N ) (N )
As soon as the transition amplitudes hkf φf |T |φi ki i for a set of positive-
energy pseudo-states are known the calculation of the ionization amplitudes
can proceed in the same way as for the CCC method, see Eq. (107). How-
ever, Zatsarinny and Bartschat [10] suggested the following modification to
the Eq. (106),
N
X
←
− (+) (−)
hΦ−
0 |H − E|Ψi i ≡ hqf |φ(N ) (N )
n ihkf φn |T |φi ki i
n=1
N
X (−)
≈ hqf |φ(N ) (N )
n ihkn φn |T |φi ki i. (118)
n=1

(−)
Here qf is a true continuum wave function of the target atom at energy
qf2 /2 = E − kf2 /2 obtained as a result of a solution of electron scattering on
the corresponding positively charged ion. Note that the replacement of kf to
kn in the T-matrix on the right-hand side of Eq. (118) produces an on-shell
T-matrix that is available from the performed scattering calculation. At this
moment we cannot provide any formal justification for this ansatz (118). We
note, however, that for the first time very good agreement with experiment
for e-He (e,2e) processes where the residual He+ ion is left in excited states
(2s+2p) was able to be obtained [10], see Sec. 4.2.4 for more detail.

3.5. Integro-differential close-coupling

The integro-differential close-coupling (IDCC) [74] represents a standard
formulation of the close-coupling method, which pioneered the application
to ionization calculations. The method starts with a multichannel expansion
of the total wave function and formulates a set of close-coupling integro-
differential equations (109) for the channel functions. These equations are
solved in coordinate space. The definition of ionization amplitude for the
e-H system that was adopted by Curran and Walters [74] can be thought of
as starting again with Eq. (47)
←− (+) (−) (−) ←
− (+)
hΦ−0 | H − E|Ψi i ≈ hq1 q2 |( H − E)I
(N )
|Ψi i
(−) (−) ←
− (N +)
= hq1 q2 |( H − E)|Ψi i. (119)
Note that the target state projection operator I (N ) appears only once on the
right-hand side of Eq. (119). This leads to an approximation of the total wave

36
 (+) (N +)
function Ψi by the close-coupling solution Ψi . The asymptotic final
(−) (−)
state is chosen as a product of two Coulomb waves q1 and q2 calculated
for charges z1 and z2 that satisfy the condition [79]
z1 z2 1 1 1
+ = + − (120)
q1 q2 q1 q2 |q1 − q2 |

The choice of the asymptotic final state is arbitrary in case of exact total
(+) (N +)
wave function Ψi , but for the approximate solution Ψi it is more crucial.
For the case of asymmetric energy sharing (q1 ≫ q2 ) considered by Curran
and Walters [74] the most important physics can be accounted for by choosing
projectile electron to be described by a plane wave (z1 = 0, k1 ≡ q1 ) and
ejected electron as a Coulomb wave (z2 = 1). Note that condition (120) is
satisfied only approximately in this case. The expression for the ionization
amplitude can now be written as
  X
(−) 1 1
hk1 q2 | − + (1 + (−1)S P12 ) |fn(N +) (r1 )φ(N )
n (r2 )i, (121)
r r12 n

(N +)
where we have substituted close-coupling expansion for the Ψi and P12
is a space exchange operator. The ionization amplitude (121) is commonly
separated into the direct and exchange terms.
With channel functions known from the solution of close-coupling equa-
tions Curran and Walters [74] have evaluated integrals in (121) numerically.
Note that for the exchange amplitude such direct evaluation of matrix ele-
ment leads to the direct overlap of two continuum functions that generally
does not converge, though Curran and Walters [74] suggested a work-around
for this problem. It is these issues, also faced by Bray et al. [75] utilising
the CCC method, that lead to the reassessment of how to apply the close-
coupling method to ionization problems [73].

4. Electron-impact ionization
4.1. Hydrogen
4.1.1. S-wave model
Before engaging in comparison with experiment it is instructive to con-
sider the consequences of using a square-integrable basis on calculating break-
up processes. The S-wave model that restricts the collision problem to just

37
 0.003 0.008
triplet singlet
CCC(100) CCC(100)
CCC(50) CCC(∞)
JS (2002) 0.007 CCC(50)
CCC(∞)
JS (2002)
0.006
singly differential cross section (a.u.)

singly differential cross section (a.u.)

0.002
0.005

0.004

0.003
0.001

0.002

0.001

0.000 0.000
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
secondary energy (a.u.) secondary energy (a.u.)

Figure 1: (Colour online) Singly differential cross sections calculated using the CCC
method with 100 and 50 Laguerre-based functions, see text. Comparison is given with
the benchmark results of Jones and Stelbovics [50]. In the singlet case, the estimates
labeled CCC(∞) are derived from the corresponding calculations, see text.

states of zero orbital angular momentum is ideal for this. In the case of e-H
scattering this is often referred to as the Temkin-Poet model due to their pro-
vision of benchmark results for the problem [39, 40]. It was originally used
to validate the convergence of the CCC method to the correct results [4] for
elastic and discrete excitation processes. It is also an ideal model for un-
derstanding how the close-coupling formalism may be applied to ionization
processes due to the availability of highly accurate results for these processes
too [50].
In figure 1 we present the singly differential cross sections (SDCS) for
40.8 eV (1.5 a.u.) electrons incident on the ground state of atomic hydrogen
(ǫi = −0.5 a.u., hence E=1 a.u.), calculated using the CCC(N) method
within the S-wave model. The conversion of the cross sections for positive-
energy states to the SDCS is given by Bray and Fursa [80]. Basically, for total
(S)
spin S the total ionization cross section σI is the integral from 0 to total

38
energy E, which is also the sum of cross sections for all open positive-energy
states. Beginning with the triplet case, the dots represent the individual cross
sections of the positive-energy states for the two Laguerre bases. We see that
there is excellent agreement with the benchmark calculations of Jones and
Stelbovics [50], but only on the interval [0, E/2]. Due to the identical nature
of the two electrons we would expect the SDCS to be symmetric about E/2 =
0.5 a.u.. However, in a unitary theory like CCC yielding symmetric SDCS
would effectively double-count the ionization processes, and so the formalism
yields near-zero cross sections for those states whose energy is greater than
E/2. In other words, the formalism attempts to enforce the integration to end
at E/2, as would be expected for identical electrons. This has a reasonable
physical interpretation. The usage of the Laguerre basis ensures that the
target-space electron is always bound and shields the scattered electron. For
(N )
target electron energies ǫn < E/2 this makes physical sense, i.e. we have
(N )
the slow electron shielding the fast one. However for ǫn > E/2 we have the
unphysical case of the fast electron shielding the slow one, and it is somewhat
satisfying that such cases yield near-zero cross sections.
Turning now to the singlet case, we see an even more interesting situation.
Rather than presenting the CCC results with dots at the secondary energies
(N )
0 ≤ ǫn ≤ E, we connect the results with lines to ensure the nature of
the oscillations is clearly visible. We see that the cross sections are non-
(N )
zero predominantly for 0 ≤ ǫn ≤ E/2, and near zero elsewhere. The
larger calculation has smaller oscillations which are around the benchmark
provided by Jones and Stelbovics [50]. When such oscillations were first
identified the suggestion was made that convergence should be to a step-
function [81]. Further analysis by Stelbovics [82] suggested that solving the
close-coupling equations for the complex amplitudes was in effect like taking
a Fourier expansion of a complex function that had a step at E/2. Owing to
the size of the step being zero in the triplet case (Pauli Principle ensures zero
cross section at E/2) the results remain smooth. In the single case, the size
of the step is non-zero and so the convergence of the amplitudes is to half
the step height, or a quarter for the cross sections. The two curves, labeled
CCC(∞) were derived from the two CCC calculations while preserving the
integral and utilising the cross section at E/2. We see very good agreement
with the results of [50]. It does not matter whether the method of solution
of the close-coupling equations is with the use of the R-matrix technique [83]
or the J-matrix method [53], the same analysis applies.

39
 The vital conclusion that we draw from the study of the S-wave model
is that the apparent conflict with formal scattering theory can be resolved.
Specifically, the physical ionization amplitudes come only from the region
(N )
of 0 ≤ ǫn ≤ E/2. Hence there is no double counting of the ionization
processes in the unitary CCC theory, though a little care needs to be taken
for the equal energy-sharing case.
All full applications of the close-coupling method to break-up processes
are affected by the above analysis. In particular, Stelbovics [82] noted that
(N )
the required ionisation amplitude FS (q, k) is defined as
(N ) (N ) (N )
FS (q, k) = fS (q, k) + (−1)S fS (k, q), (122)
(N )
but only if fS (q, k) (see Eq. (107)) is zero for q > k, i.e. if the step function
is satisfied to sufficient numerical precision. Subsequently, for k = q it was
(N ) (N )
found that fS (q, k) ≈ (−1)S fS (k, q) holds, so that
 
(N ) (N ) (N ) (N )
2 |fS (q, k)|2 + |fS (k, q)|2 ≈ |fS (q, k) + (−1)S fS (k, q)|2 , (123)

see Eq. (15) of Bray [84]. Thus in CCC calculations the equal energy-sharing
amplitudes arise entirely ab initio. However, the asymmetric amplitudes will
show minor oscillations that may be eliminated with the aid of a smooth
N = ∞ estimate of the SDCS, as was obtained in Fig. 1.
Before we look at fully differential cross sections at various energies it is
important to establish the accuracy of the method for the total ionization
cross section at all energies of interest. This is presented in Fig. 2. It was
the outstanding agreement between the CCC theory and experiment for total
ionization, without any double-counting problems, that lead to our extension
of the CCC method to fully differential ionization processes.

4.1.2. Low energies

We now turn to full calculations of e-H ionization, starting with the low
energies, while restricting our discussion to primarily fully differential cross
sections. Atomic hydrogen is the ideal testing ground for electron-impact ex-
citation and ionization theories. Unfortunately, the experimental difficulties
of creating a beam of atomic hydrogen suitable for collision studies are quite
substantial, as is putting the measured data onto an absolute scale [86]. This
typically leads to larger experimental uncertainties for e-H ionisation as com-
pared to say to the e-He case, and is particularly a problem at low energies
where the cross sections are small.

40
 7
exp, JPB (1987)
6 CCC, PRL (1993)
cross section (10-17cm2)

0
1 10 100 1000
total energy E (eV)

Figure 2: Total electron-impact ionization cross section of the ground state of atomic
hydrogen. Experiment is due to Shah et al. [85] and the CCC results are from Bray and
Stelbovics [43].

As far as we are aware detailed fully differential cross sections, whose

geometries are summarized in figure 3, for e-H ionization are only available
for incident energies of 15.6 eV [86, 87], 17.6 eV [87, 88], 20 eV, 25 eV,
30 eV [89, 90], 27.2 eV [91], 54.4 eV [92] and 150 eV [93]. The first five
energies have data for equal energy-sharing, while the latter two, as well as
at 27.2 eV, have asymmetric energy sharing data. Only the data at 15.6 eV,
17.6 eV, 54.4 eV and 150 eV have been put on the absolute scale. All data
are for the coplanar geometry (ψ = 0◦ in figure 3). Note that doubly differ-
ential ionization cross sections (the equivalent of excitation differential cross
sections) are also available [94].
There have been many theories applied to fully differential ionization
at the lowest energy case1 of 15.6 eV. Though one would expect only non-
perturbative approaches to work well here we have some first order theories

1
There is the single geometry of θAB = 180◦ available at 14.6 eV [86]

41
 E inc

EB

ψ θB
θA

EA

Figure 3: Parametrization of electron-impact single ionization fully differential cross sec-

tions. An electron of energy Einc is incident at an angle ψ relative to the plane formed by
the two outgoing electrons of energies EA and EB .

capturing much of the electron-electron correlations, see for example Jones

et al. [95] and references therein. However, a fully quantitative agreement
can only be obtained utilising non-perturbative approaches such as ECS,
TDCC, and CCC. The excellent agreement between these three theories in
both shape and magnitude is demonstrated in Fig. 4. Though we note that
the demonstrated quantitative agreement is only possible when experimental
mean values, which have an uncertainty of 35%, are divided by two. This
requirement was identified quite early, even when the full understanding of
the CCC method had not yet emerged [96].
It is quite remarkable to see such good agreement between the ECS,
TDCC and CCC calculations for all of the geometries presented in figure 4.
There are three distinct geometries presented which are slices in the three-
dimensional surface that is a function of the scattering angles θA and θB .
The symmetric geometry (θA = −θB ) has the detectors on the opposite sides
of the incident beam. It is particularly helpful in checking the accuracy of
the CCC calculations by checking how close to zero the cross sections are
when the two electrons go out together in the forward or backward angles.
Such cross sections can only come out to be small through the destructive
interference of the underlying partial-wave scattering amplitudes. This re-
quires sufficiently large target-space orbital angular momentum l. Presently,
we took l ≤ lmax = 6, with Nl = 70 − l and λl = 2, while keeping all
open and a few closed states. The fixed-θA geometries provide for a set of

42
 15 20
θA=-θB θAB=100o
15
10 exp/2
ECS 10
5 TDCC
CCC
5
0 0
0 30 60 90 120 150 180 -180 -120 -60 0 60 120 180
200 75

cross section (10-20cm2sr-2eV-1)

θA=-30o θAB=120o
150
50
100
25
50
0 0
-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180
20 250
θA=-90o 200 θAB=150o
15
150
10
100
5 50
0 0
-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180
150 300
θA=-150o 250 θAB=180o
100 200
150
50 100
50
0 0
-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180
scattering angle θB (deg) scattering angle θB (deg)

Figure 4: Coplanar fully differential cross sections for 15.6 eV electrons incident on the
ground state of atomic hydrogen with 1 eV outgoing electrons. The exterior complex
scaling (ECS) results are due to Baertschy et al. [2]. The time-dependent close-coupling
(TDCC) results are due to Colgan and Pindzola [11]. The Laguerre-based CCC calcula-
tions are described in the text. The absolute experimental values, with an uncertainty of
35%, due to Röder et al. [86], have been divided by two for best visual fit to theory.

systematic slices through the cross section surface. As with the symmetric
geometry, we expect near zero cross sections whenever θA ≈ θB . Lastly, the
fixed-θAB = θB − θA geometries are very useful for seeing quantitatively the
effect of electron-electron correlation in the final state. For the smaller θAB
we expect generally smaller cross sections. The maximal values of the cross
sections arise in the Wannier geometry of θAB = 180◦. Note how the maxima
only arise when one of the electrons is scattered into the forward direction.
It is our contention that the 15.6 eV case is sufficient to demonstrate
that the ECS, TDCC, and CCC theories are able to solve the e-H ionization
problem at any energy. However, having to divide the experimental values
by two has to be of some concern, even when the stated absolute uncertainty
is 35%. We therefore turn to the 17.6 eV case, the next energy at which
absolute data are available. In this case there was some problem with in-
ternal consistency of the measured data, but this was resolved in the final
presentation by Röder et al. [88]. Colgan and Pindzola [11] have already
shown excellent agreement between the ECS, TDCC and CCC calculations
at 17.6 eV. Following Röder et al. [88], they did not scale experiment by any
factor, and found the theoretical results systematically a little lower than
experiment. Keeping in mind that the absolute uncertainty is 40% [86], in
figure 5 we divide the experiment by 1.4 for best overall visual fit to theory.

43
 90 70
14 θA = -θB 80 θ = 140o 60 θAB = 120
o
12 A
exp/1.4 70
10 60 50
ECS
8 TDCC 50 40
6 CCC 40 30
30 20
4 20
2 10 10
0 0 0
cross section (10-20cm2sr-2eV-1)

0 30 60 90 120 150 180 -180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180

160 12 160
140 θA = -20o 10 θAB = 90
o 140 θAB = 150o
120 120
100 8 100
80 6 80
60 4 60
40 40
20 2 20
0 0 0
-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180

30 25 250
o
25 θA = 60 θAB = 100o θAB = 180o
20 200
20
15 150
15
10 100
10
5 5 50
0 0 0
-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180
scattering angle θB (deg)

Figure 5: Coplanar fully differential cross sections for 17.6 eV electrons incident on the
ground state of atomic hydrogen with 2 eV outgoing electrons. The ECS and TDCC results
are as for Fig. 4. The CCC calculation is described in the text. The absolute experimental
values, with an uncertainty of 40%, due to Röder et al. [88], have been divided by 1.4 for
best visual fit to theory.

The present CCC calculations supersede those presented by Röder et al.

[88]. We increased lmax from 5 to 7, while leaving Nl = 50 − l and λl = 2.
The effect of the increased lmax was to reduce the cross section at the forward
scattering angles of the θA = −θB geometry. The agreement demonstrated
in figure 5 between theory and experiment is quite remarkable, and gives us
further confidence that the ECS, TDCC and CCC theories are able to fully
solve such ionization problems.

4.1.3. Intermediate energies

The 20, 25, and 30 eV electron-impact ionization data [89, 90] is for equal
energy-sharing, and remains on a relative scale, but is internomalised across
the various geometries, but not energy.
It is very instructive to compare the Laguerre-based (CCC-L) and Box-
based (CCC-B) approaches to electron-atom ionization collision problems.

44
In the former, the extent of the radial functions depends on the combination
of Laguerre basis Nl , the Laguerre exponential fall-off parameter λl , and
the energy of the state. Roughly, the number of oscillations in all positive-
energy states is the same, with the lower energy states therefore extending
further in coordinate space. However, in CCC-B all positive energy states
stop at r = R0 , with higher energy states having many more oscillations
than the lower energy ones. The usage of λl = λ allows for a systematic
convergence study [97]. It was the varying of λl to avoid interpolation across
the complex amplitudes that resulted in some of the discrepancies in the
earlier applications of the CCC-L method [84].
One advantage of the CCC-B approach is that the positive target-state
energies are linear in momentum, whereas they grow almost exponentially
for CCC-L. This means that in the case of equal-energy sharing cases at
the higher energies, CCC-B is a little more advantages by providing a more
dense discretization around the required energy point. Accordingly, in Fig. 6
the CCC calculations are all performed in the CCC-B mode with all having
lmax = 7, Nl = 25 − l, but with R0 =90, 70, and 60 a.u. for the incident
electron energies of 20, 25, and 30 eV, respectively. Comparison is given with
experiment and the ECS calculations, with outstanding agreement found in
all cases.

4.1.4. High energies

At the higher energies of 54.4 and 150 eV the experimental data is ab-
solute, and is for asymmetric energy sharing. These two cases were the first
to be considered when developing the CCC approach to ionization [75]. At
that time we followed the approach of Curran and Walters [74], but ran
into problems with non-existent integrals. The present implementation is
much simpler, having no reconstruction of the total wavefunction, and de-
rives the ionization amplitudes directly from the excitation amplitudes of the
positive-energy pseudostates. At the time the 54.4 eV data was not available
on the absolute scale, but it is now [98], and is presented in Fig. 7. Excellent
agreement is found between experiment and the CCC calculations after the
former have been uniformely divided by 0.8, which is well withing the ex-
perimental uncertainty of 35%. For interest, we ran the CCC calculations in
the distorted-wave Born approximation (DWBA) mode (coupling excluded)
which gives a qualitative account of the experimental observations. The same
distorting potential (ground state) is used for the initial and final projectile
space electrons, with the ejected electron treated via the positive-energy pseu-

45
 5 5 8
20 eV θA = -θB 25 eV θA = -θB 7 30 eV θA = -θB
4 4 6
3 3 5
4
2 2 3
1 1 2
1
0 0 0
0 30 60 90 120 150 180 0 30 60 90 120 150 180 0 30 60 90 120 150 180
9 18 25
8 θAB = 80o 16 θAB = 80o θAB = 80o
7 14 20
6 12
5 10 15
4 8 10
3 6
2 4 5
1 2
0 0 0
-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180
16 25 30
14 θAB = 90o θAB = 90o 25 θAB = 90o
12 20
10 20
15
8 15
6 10
10
4 5
2 5
0 0 0
cross section (10-20cm2sr-2eV-1)

-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180
30 30 30
25 θAB = 100o 25 θAB = 100o 25 θAB = 100o
20 20 20
15 15 15
10 10 10
5 5 5
0 0 0
-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180
60 45 30
50 θAB = 120o 40 θAB = 120o 25 θAB = 120o
35
40 30 20
30 25 15
20
20 15 10
10 10 5
5
0 0 0
-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180
120 70 40
100 θAB = 150o 60 θAB = 150o 35 θAB = 150o
50 30
80 25
40
60 20
30 15
40 20 10
20 10 5
0 0 0
-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180
160 80 40
140 θAB = 180o 70 θAB = 180o 35 θAB = 180o
120 60 30
ECS
100 50 25 CCC
80 40 20
60 30 15
40 20 10
20 10 5
0 0 0
-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180
scattering angle θB (deg)

Figure 6: Coplanar fully differential cross sections for 20 eV (left), 25 eV (center) and
30 eV (right) electrons incident on the ground state of atomic hydrogen with equal energy
outgoing electrons. The ECS theory is due to Baertschy et al. [2]. The CCC calculation
is described in the text. The relative experimental values, due to Röder et al. [89] and
Whelan et al. [90], are internormalised at each energy across the various geometries.

46
 300 180
θA = 4o 160 θA = 16o
250
140
200 120
100
150
80
cross section (10-20cm2sr-2eV-1)

100 60
40
50
20
0 0
-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180

250 100
θA = 10o 90 θA = 23o
200 80 exp/0.8
70 CCC
DWBA
150 60
50
100 40
30
50 20
10
0 0
-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180
scattering angle θB (deg)

Figure 7: Coplanar fully differential cross sections for 54.4 eV electrons incident on the
ground state of atomic hydrogen with an outgoing electron of energy EB =5 eV. The
fast electron of energy EA =35.8 eV is detected at the specified θA angles. The relative
experimental data presented by Brauner et al. [92], have been put on the absolute scale
with an uncertainty of 35% by Jochen Röder [98], and divided by 0.8 for best visual fit to
the CCC theory. The CCC and DWBA calculations are described in the text.

dostates. Note that the DWBA results typically vary with the choice of the
distorting potential, whereas the CCC results must be independent of such
a choice.
The 150 eV case is presented in Fig. 8 for completeness. This has been
a solved problem for a very long time, and we see that the DWBA and
CCC results are barely distinguishable. Here we took the opportunity to
update the CCC results using the present CCC approach to ionization that
superceded the one originally applied [75]. We set lmax = 6 and Nl = 25 − l
with λl = 2. The DWBA mode turns off exchange and coupling.
We conclude this section by supporting the comments of Rescigno et al.
[1] that the electron-impact ionization problem has been reduced to simply a
computational one. Presently, we have the ECS, TDCC and CCC methods
that are all able to solve such problems. Interestingly, computational progress

47
 100 70 30
90 3 eV θA = 4o 60 5 eV θA = 4o 25 10 eV θA = 4o
80
70 50 20
60 40
50 15
40 30
30 20 10
20 5
10 10
0 0 0
cross section (10-19cm2sr-2eV-1)

-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180
40 35 25
35 θA = 10o 30 θA = 10o θA = 10o
30 20
25
25 15
20
20
15 10
15
10 10
5
5 5
0 0 0
-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180
10 12 16
9 θA = 16o 10 θA = 16o 14 θA = 16o
8 12
7 8 exp/1.2
6 10 CCC
5 6 8 DWBA
4 6
3 4
2 4
2 2
1
0 0 0
-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180
scattering angle θB (deg)

Figure 8: Coplanar fully differential cross sections for 150 eV electrons incident on the
ground state of atomic hydrogen with outgoing electrons of energy EB = 3 eV (left), 5 eV
(center) and 10 eV (right). The fast electron is detected at the specified θA angles. text.
The absolute experimental data, due to Ehrhardt et al. [93], has been divided by 1.2 for
best visual fit to theory. The CCC and DWBA calculations are described in the text.

was possible without a full understanding of the origins of something as

fundamental as the ionization amplitude [15].

4.2. Helium
Whereas atomic hydrogen is the ideal target for theorists, helium is the
ideal target for experimentalists whilst maintaining the theoretical need for
simplicity. It turns out that the helium atom is very well described by
the frozen-core model, where one of the electrons occupies the He+ (1s) or-
bital [99]. Thus, the appropriately symmetrised configurations are of the
form {1s, nl}. The greatest error is in the ionization energy of the ground
{1s, 1s′} state, being 23.74 eV rather than the experimental value of 24.58 eV.
The CCC method for helium within the frozen-core model has been thor-
oughly tested for electron-impact excitation of n ≤ 3 singlet and triplet
states [99, 100]. Nevertheless, due to the structural approximations, we can-

48
 40

35
cross section (10-18cm2)

30

25

20

15

10

5 Rejoub et al
Sorokin et al
CCC
0
100 1000
projectile energy (eV)

Figure 9: The total electron-impact single ionization cross section of the ground state of
helium. The experiments are due to Rejoub et al. [105] and Sorokin et al. [106], and the
CCC theory from Bray and Fursa [102].

not expect the same level of precision as for atomic hydrogen. The extension
to electron-impact ionization processes for helium [73] follows in the same
way as for atomic hydrogen, though the treatment of the equal-energy shar-
ing case required more sophistication [101].
As for atomic hydrogen we begin by examining the total electron-impact
(single) ionization cross section of helium, given in Fig. 9. Here we present
the comparison between the multi-configurational structure model for helium
and the more recent experiments. The difference between the frozen-core and
multi-configuration results is relatively small and has already been discussed
in some detail [102]. The key point is that as for atomic hydrogen the CCC
theory yields excellent agreement with experiment across all energies of in-
terest. As far as we are aware it remains the only theory to date that is able
to do so. The ECS method has not yet been extended to full e-He prob-
lems, though some progress in this direction has been made [60]. The TDCC
implementation for helium [103, 104] uses the configuration average approx-

49
 15 10 15
θA=−θB θA=120o θAB=135o
Rösel et al.
10 10
5
5 5

0 0 0
0 30 60 90 120 150 180 −180 −120 −60 0 60 120 180 −180 −120 −60 0 60 120 180
20 15 20
cross section (10−20cm2eV−1 sr−2)

θA=30o θA=140o θAB=150o

15 15
CCC−L 10
10 CCC−B 10
5
5 5
0 0 0
−180 −120 −60 0 60 120 180 −180 −120 −60 0 60 120 180 −180 −120 −60 0 60 120 180
10 20 30
θA=60o θA=150o 25 θAB=165o
15
Jones at al. 20
5 10 15
10
5
5
0 0 0
−180 −120 −60 0 60 120 180 −180 −120 −60 0 60 120 180 −180 −120 −60 0 60 120 180
10 10 30
θA=90o θAB=120o 25 θAB=180o
20
5 5 15
10
5
0 0 0
−180 −120 −60 0 60 120 180 −180 −120 −60 0 60 120 180 −180 −120 −60 0 60 120 180
scattering angle θB (deg)

Figure 10: Fully differential cross sections for 26.6 eV electron-impact ionization of the
ground state of helium with 1 eV outgoing electrons in the coplanar geometry. The
experiment, with an absolute uncertainty of 22%, and the Jones et al. calculations are
presented by Rösel et al. [107]. The CCC calculations are due to Stelbovics et al. [101].
Note, wherever possible points from overlapping geometries are also plotted, giving an
indication of internal consistency.

imation, and so has difficulty at low energies when electron flux needs to be
correctly subdivided between the singlet and triplet states. An all-electron
TDCC implementation [65] should be able to do this, but it is particularly
computationally expensive. RMPS approach should work well, but has con-
centrated on ionization-plus-excitation processes [10].
The electron-impact fully differential single ionization problem is arguably
the most studied for the helium target than any other, both experimentally
and theoretically. Accordingly, it is not practical to review all of the available
studies, and we choose some specific highlights. It is our contention that this
problem is as solved as it is for atomic hydrogen, though some few nagging
discrepancies with experiment remain.

50
4.2.1. Low energies
The lowest energy for which e-He fully differential cross sections exist
is 26.6 eV. In this case coplanar kinematics with 1 eV outgoing electrons
were chosen. This is the most difficult case to calculate due to the very low
energies involved with the competing elastic and excitation cross sections
being considerably larger. The comparison between the Laguerre- and Box-
based CCC theory and the experiment and calculations of Rösel et al. [107]
are presented in Fig. 10. We see vey good agreement between the two CCC
calculations and the experiment. Remarkably, the first order calculations of
Jones et al are only in marginally worse agreement with experiment.
As in the case of atomic hydrogen it is tempting to conclude that the e-He
single ionization problem is also solved for all practical purposes. Particularly
so since similarly good agreement between the CCC theory and experiment
is also found for 32.6 eV incident energy with 4 eV outgoing electrons [101].
However, thus far we have only looked at the coplanar case. Out-of-plane
measurement are only available for the intermediate and high energy cases.

4.2.2. Intermediate energies

We’ll define, somewhat aribitrarily, the intermediate energy range of in-
terest to be above 32.6 eV and below 100 eV. Here data are available for
both equal- and unequal-energy-sharing kinematics. Additionally, data for
out-of-plane geometries are also available. While we will restrict ourselves
to the geometries conveniently parametrised as in Fig. 3, there are others as
well, see Röder et al. [108] for example of good agreement between theory
and experiment at 40 eV, and 44.6 eV [109]. Note that in the latter case the
factor of two difference was derived from first principles subsequently [101].
We’ll begin by revisiting the 50 eV e-He case (E = 25.4 eV) first consid-
ered by Röder et al. [111]. Here, we have three energy-sharings measured for
EB = 4 eV, EB = 10 eV, and EA = EB = 12.7 eV. Only the data for the
asymmetric energy-sharing cases has been put on the absolute scale. This
case is interesting because it relates to the step-function issues discussed ear-
lier in Section 4.1.1. We begin by considering the singly differential cross
section (SDCS), presented in Fig. 11. Here we have the results of a lmax = 6,
50 a.u. box-based 283-state CCC calculation which shows oscillations for
EB ≤ E/2, and falls off smoothly to zero for EB > E/2. The curve labeled
CCC(∞) is the corresponding step function estimate. It preseves the integral
of the CCC(283) curve, that yields the kind of agreement with experiment
presented in Fig. 9, with the height of the step being four times the presented

51
 4

cross section (10-18cm2eV-1)

3

50eV e-He SDCS

exp
CCC(283)
CCC(∞)
0
0 5 10 15 20 25
secondary energy EB (eV)

Figure 11: The 50 eV electron-impact on the ground state of helium singly differential cross
section. The experiment is due to Röder et al. [110], and the present CCC calculations
are described in the text.

CCC(283) result. The coherent combination of the ionization amplitudes at

EB = E/2 yields cross sections of the correct magnitude ab initio [101], but
for the oscillations for EB ≤ E/2 invite some further attention, particularly
for EB = 10 eV. In this case we can be sure that the raw CCC(283) fully
differential cross sections, will be substantially too low upon integration over
the angular coordinates. Accordingly, we multiply them uniformly by the
ratio of the CCC(∞) and CCC(283) results for each EB < E/2 of interest.
The resultant cross sections and comparison with experiment are presented
in Fig. 12.
We see excellent agreement with experiment, which is of considerable
improvement over what was presented earlier [111]. In part this is due to
the much larger calculations now possible, but also due to the greater un-
derstanding of how the close-coupling method should be applied to ionizing
processes.
Thus far we have concentrated on cases where there is clearly good agree-
ment between theory and experiment. However, this is not always the case.
In Fig. 13 we summarise an extraordinary situation that exists for 64.6 eV
electron-impact ionisation of the ground state of helium with equal energy
(20 eV) outgoing electrons. This case has been studied in great detail al-
ready [113] using the CCC theory and TDCC [114, 115] methods. Here we

52
 35 8 8
30 EB=4eV, θA=20o 7 EB=4eV, θA=35o 7 EB=10eV, θA=30o
25 exp 6 6
CCC 5 5
20
4 4
15
3 3
10 2 2
cross section (10-20cm2eV-1 sr-2)

5 1 1
0 0 0
-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180
20 14 6
EB=4eV, θA=25o θ EB=10eV, θA=35o
o
18 EB=10eV, A=20
12 5
16
14 10
4
12 8
10 3
8 6
6 2
4
4 1
2
2
0 0 0
-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180

14 10 3
EB=4eV, θA=30o 9 EB=10eV, θA=25o EA=EB, θA=-θB
12
8
10 7 2
8 6
5
6 4
3 1
4
2
2 1
0 0 0
-180 -120 -60 0 60 120 180 -180 -120 -60 0 60 120 180 0 30 60 90 120 150 180
scattering angle θB (deg)

Figure 12: The 50 eV electron-impact on the ground state of helium coplanar fully dif-
ferential cross sections at the specified secondary energies and geometries. The absolute
experiment at EB = 4, 10 eV is due to Röder et al. [111], while the relative equal-energy
measurements of Rösel et al. [112] have been normalised to theory. The present CCC
calculations are described in the text.

highlight two independent issues using just the CCC theory for clarity of
presentation.
Begining with the perpendicular plane case (ψ = 90◦ ), we see outstanding
agreement between theory and experiment. The fact that the three geome-
tries have the θA = 90◦ point in common (see Fig. 3) means that the full
data set is internormalised, and so we are not free to normalise each case
separately. If we are to believe that the CCC theory is correct for ψ = 90◦
that would suggest that the absolute uncertainty is of the order of 0.01 of
the specified units. However, the much larger coplanar case (ψ = 0◦ ) the
discrepancy arond θA = 40◦ is 200 times bigger! The situation would be
resolved to a substantial extent if we were allowed to normalise the two cases
separately, but the small error bars do not allow for the required movement.

53
 6.00 0.45 0.18
ψ=0o ψ=68o ψ=90o
cross section (10-20cm2sr-2eV-1)

exp 0.40 0.16

5.00 CCC
0.35 0.14

4.00 0.30 0.12

0.25 0.10
3.00
0.20
0.08
0.15
2.00 0.06
0.10
0.04
1.00 0.05
0.02
0.00
0.00 0 30 60 90 120 150 180 0.00
0 30 60 90 120 150 180 scattering angle θA=-θB (deg) 0 30 60 90 120 150 180

Figure 13: The 64.6 eV electron-impact on the ground state of helium fully differential
cross sections with 20 eV outgoing electrons for the geometries as specified in Fig. 3. The
relative experimental data [116] has been normalised to the CCC calculations [113] at the
θA = 90◦ point, which is common for all ψ.

The second issue we’d like to bring to the attention of the reader is the
ψ = 68◦ case, where the cross section appears to go to zero at θA ≈ 70◦ .
As has been extensively studied by Colgan et al. [115], this minimum comes
about from a destructive interference of the contributing partial waves. We
found that taking ψ = 62.7◦ the minimum was five orders of magnitude less
than the maximum of the cross section. By excluding contributions from
the larger orbital angular momenta the minimum rises rapidly. Accordingly,
there is no readily identifiable mechanism that is responsible for this, and it
is conceivable that the minimum is an actual zero.
We next turn to one of the most spectacular experimental achievements
in the field in recent times. This is the application of the cold-target re-
coilion momentum spectroscopy (COLTRIMS) technique, for a given inci-
dent energy, to measure absolute fully differential ionization cross sections
simultaneously for many kinematical and geometrical combinations of the
outgoing electrons. Given the CCC cross section behaviour identified in
Fig. 1, this is ideal in testing the CCC theory as the secondary energy EB
varies in small steps from the minimum (EB = 3 eV) to the maximum value
(EB = EA = 23 eV) measurable. Comparison of experiment with the La-
guerre based CCC results in the coplanar case is presented in Fig. 14. We

54
 120 70 30 16 8
100 EB=3eV θΑ=-10o 60 θΑ=-15o 25 θΑ=-20o 14 θΑ=-25o 7 θΑ=-30o
50 12 6
80 20 10 5
40
60 15 8 4
30 6 3
40 20 10
4 2
20 10 5 2 1
0 0 0 0 0
0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360
90 45 30 16 9
80 EB=5eV θΑ=-10o 40 θΑ=-15o 25 θΑ=-20o 14 θΑ=-25o 8 θΑ=-30o
70 35 12 7
60 30 20 10 6
50 25 5
15 8
40 20 4
30 15 10 6 3
20 10 4 2
5 2
10 5 1
0 0 0 0 0
0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360

60 35 25 18 10
EB=8eV θΑ=-10o 30 θΑ=-15o θΑ=-20o 16 θΑ=-25o 9 θΑ=-30o
50 20 8
14
40 25 12 7
20 15 10 6
30 5
15 10 8 4
20 10 6 3
cross section (10-20cm2sr-2eV-1)

5 4 2
10 5 2 1
0 0 0 0 0
0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360
40 30 25 16 10
35 EB=10eV θΑ=-10o θΑ=-15o θΑ=-20o 14 θΑ=-25o 9 θΑ=-30o
25 20 8
30 12
20 7
25 15 10 6
20 15 8 5
15 10 6 4
10 3
10 4
5 5 2
5 2 1
0 0 0 0 0
0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360
30 20 18 14 10
EB=12eV θΑ=-10o 18 θΑ=-15o 16 θΑ=-20o 12 θΑ=-25o 9 θΑ=-30o
25 16 8
14
20 14 12 10 7
12 10 8 6
15 10 5
8 8 6 4
10 6 6 4 3
4 4 2
5 2
2 2 1
0 0 0 0 0
0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360

20 16 16 14 10
EB=15eV θΑ=-10o 14 θΑ=-15o 14 θΑ=-20o 12 θΑ=-25o 9 θΑ=-30o
15 12 12 8
10 7
10 10 8 6
10 8 8 5
6 6 6 4
5 4 4 4 3
2
2 2 2 1
0 0 0 0 0
0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360
10 6 8 6 4.5
EB=23eV θΑ=-25o 5 θΑ=-45o 7 θΑ=-60o 5 θΑ=-75o 4 θΑ=-122o
8 3.5
6
4 5 4 3
6 2.5
3 4 3
4 2
2 3 2 1.5
2 1
2 1 1
1 0.5
0 0 0 0 0
0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360
scattering angle θB (deg)

Figure 14: The 70.6 eV electron-impact on the ground state of helium fully differential
cross sections for the coplanar geometry. The absolute experiment and CCC theory are
due to Ren et al. [104].

55
see generally outstanding agreement across 35 separate cases, all of which
arise from a single experiment and a single calculation. Incidentally, the
TDCC results are also presented by Ren et al. [104], but they suffer from a
systematic discrepancy with experiment for the asymmetric energy sharing
cases.
In addition to the coplanar data Ren et al. [104] also extracted data
where one of the electrons is ejected perpendicular to the plane formed by
the incident and the so-called scattered electron. Given the discrepancy
identified at 64.6 eV in Fig. 13 it is helful to see if the CCC theory is able
to yield consistent agreement with absolute experiment for both coplanar
and perpendicular-plane geometries. The perpendicular case is presented
in Fig. 15. Once more we see generally excellent agreement of the CCC
calculations with experiment, particularly given the experimental energy and
angle resolution considerations. It is important to remember that being a
unitary theory the T -matrix elements that arise from a single calculation are
all coupled, and thereby contribute in some way to all of the presented cross
sections. Though not conclusive, the weight of comparison between the CCC
theory and experiment at 70.6 eV suggests similarly good agreement should
be expected at 64.6 eV.

4.2.3. High energies

For completeness we say a few words about the higher energies. As for
atomic hydrogen such data have been the most studied and is the fundamen-
tal starting point for the testing of any theory, including CCC [73]. Here
perturbative methods are a very efficient way of yielding accurate cross sec-
tions. Hence, the intracacies of the definition of the ionization amplitude are
not so relevant here.
Most of the experimental data is several decades old as are the calcu-
lations, but there have been a few more recent developments. Dürr et al.
[117] have measured 102 eV electron-impact ionization of helium using the
COLTRIMS techniques and found good agreement with various theories.
This case was also studied with some success using the TDCC theory [66].
Catoire et al. [118] measured 730 eV electron-impact ionization of helium
with a coplanar ejection of a 205 eV electron, and found some disagreement
with various perturbative theories. This improved substantially upon appli-
cation of the CCC theory [119], however being relative measurements the
comparison is not able to be definitive.

56
 120 50 25 10 6
EB=3eV θΑ=-10o 45 θΑ=-15o θΑ=-20o 9 θΑ=-25o θΑ=-30o
100 40 20 8 5
80 35 7 4
30 15 6
60 25 5 3
20 10 4
40 15 3 2
20 10 5 2 1
5 1
0 0 0 0 0
0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360
80 40 18 10 6
70 EB=5eV θΑ=-10o 35 θΑ=-15o 16 θΑ=-20o 9 θΑ=-25o θΑ=-30o
8 5
60 30 14
12 7 4
50 25 6
10
40 20 5 3
8 4
30 15 6 2
3
20 10 4 2 1
10 5 2 1
0 0 0 0 0
0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360

50 30 16 9 4.5
45 EB=8eV θΑ=-10o θΑ=-15o 14 θΑ=-20o 8 θΑ=-25o 4 θΑ=-30o
40 25
12 7 3.5
35 20 6 3
30 10
5 2.5
25 15 8
20 4 2
10 6 3 1.5
15 4
10 2 1
5 2 1 0.5
5
0 0 0 0 0
0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360
35 20 12 7 4
30 EB=10eV θΑ=-10o 18 θΑ=-15o θΑ=-20o 6 θΑ=-25o 3.5 θΑ=-30o
16 10
25 5 3
14 8
12 2.5
20 4
10 6 2
15 8 3
4 1.5
10 6 2 1
4 2
5 2 1 0.5
0 0 0 0 0
0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360
25 16 10 6 4
EB=12eV θΑ=-10o 14 θΑ=-15o 9 θΑ=-20o θΑ=-25o 3.5 θΑ=-30o
20 8 5
12 3
7 4
15 10 6 2.5
8 5 3 2
10 6 4 1.5
3 2
4 1
5 2 1
2 1 0.5
0 0 0 0 0
0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360

16 14 8 5 4
14 EB=15eV θΑ=-10o 12 θΑ=-15o 7 θΑ=-20o 4.5 θΑ=-25o 3.5 θΑ=-30o
12 6 4 3
10 3.5
10 8 5 3 2.5
8 4 2.5 2
6 6 3 2 1.5
4 4 2 1.5 1
1
2 2 1 0.5 0.5
0 0 0 0 0
0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360
4 2.5 3.5 3.5 4.5
3.5 EB=23eV θΑ=-25o θΑ=-45o 3 θΑ=-60o 3 θΑ=-75o 4 θΑ=-122o
2 3.5
3 2.5 2.5
2.5 3
1.5 2 2 2.5
2
1 1.5 1.5 2
1.5 1.5
1 1 1
0.5 1
0.5 0.5 0.5 0.5
0 0 0 0 0
0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360 0 60 120 180 240 300 360

Figure 15: The 70.6 eV electron-impact on the ground state of helium fully differential
cross sections for the perpendicular plane geometry. The absolute experiment and CCC
theory are due to Ren et al. [104].

57
4.2.4. Ionization plus excitation
Thus far we have considered electron-impact ionization of helium in the
case where the residual electron is left in the ground state of the He+ ion.
The formal theory for the definition of the ionization amplitude is unchanged
for the case where the residual ion is left in an excited nf ≥ 2 state. How-
ever, perhaps surprisingly, the computation of such processes is not a trivial
extension of what has been presented thus far. In a close-coupling based
theory the same definition (106) for the ionization amplitude applies
N
X
(N ) (−) (N ) ←
− (N +)
fi (qf , kf ) = hqf |φ(N )
n ihkf φn |( H − E)|Ψi i, (124)
n=1

where now hqf | is a continuum wave obtained as a close-coupling solution of

e-He+ scattering that includes at least the lowest three energy states, and
(N )
the helium states |φn i are constructed from configurations that contain at
least the lowest three states of He+ for one of the electrons. One consequence
(−) (N ) (−) (N )
of this is that it is no longer possible to obtain hqf |φn i = δf n hqf |φn i
(N ) ←
− (N +)
for any qf2 /2. Since hkf φn |( H − E)|Ψi i exists only on the energy shell
2 (N )
(kf /2 + ǫn = E), we have a fundamental computational problem.
There are several hybrid approaches to ionization plus excitation pro-
cesses, see [120–124] for example. However, as stated at the outset, here
we only concern ourselves with those approaches that aim to fully solve the
problem, i.e. obtain an accurate ionization amplitude at any energy. As
far as we are aware only Zatsarinny and Bartschat [10] have attempted to
do this. Utilising their close-coupling formalism they effectively replaced kf
with kn on the RHS of Eq. (124), and thereby avoided the above-mentioned
problem. This means that the summation brings contributions from excita-
tion amplitudes with rather different kinematics than in the experiment of
(−) (N )
interest, depending on the magnitudes of the hqf |φn i. Nevertheless, the
results, reproduced here in Fig. 16, are quite spectacular. We see very good
agreement of the B-spline R-matrix (BSR) calculations with the ratio mea-
surements of Bellm et al. [124]. At times the “nf = 1” cross section is two
orders of magnitude bigger than the “nf = 2” one. This indicates that the
absolute “nf = 2” cross sections are being calculated very accurately even
though they are particularly small.
As this is only a very recent development more studies are required to
determine the validity of the approach at other energies, and for the residual
ion being in He+ (nf > 2) states.

58
5. Double photoionization
The process of non-sequential single-photon double photoionization (DPI),
and a closely related process of photoionization-excitation, both involve ab-
sorption of a single photon which results in simultaneous change of quantum
numbers of two or more atomic electrons. These are fundamentally impor-
tant processes as they constitute a direct probe of electron correlation [125].
In the independent electron approximation, at most practical photon ener-
gies, the electromagnetic field can only couple to a single electron. Changing
the quantum state of the second and further electrons can only take place via
many-electron correlation, either in the initial bound state of the target or the
final multiply ionized state. The bound state correlation is a static process
which can be described relatively easily by various configuration-interaction
schemes. Dynamical correlation in the photoelectron continuum is much
more difficult to describe theoretically. Due to a long range of the Coulomb
interaction, the continuum correlation cannot be treated perturbatively and
a non-perturbative approach has to be taken.
A two-electron continuum state can be treated very efficiently by the con-
vergent close-coupling (CCC) method in which one of the electrons is repre-
sented by a complete set of discrete (positive and negative energy) Laguerre
basis pseudostates while the second electron is treated as a true continuum
state. Details of the CCC method, as applied to electron-atom scattering,
are outlined in the previous chapters.

5.1. Helium
Application of the CCC method do DPI of the helium atom is straight-
forward [126]. After the photon is absorbed and the first electron is ejected
into the continuum, the atomic system evolves through elastic and inelastic
scattering of the photoelectron on the He+ ion. This process can be described
by the multi-channel expansion for the final state wave function:
XX Z
(−) (−) hp(+) i|T |jk(−) i
|Ψj (k)i = |jkb i + d3 p 2 /2 − ǫ + i0
|ik(+) i , (125)
i
E − p i

with boundary conditions corresponding to an outgoing wave in a given chan-

nel |jk(−) i and incoming waves in all other channels |ip(+) i. The total inte-
grated cross section (TICS) of the helium photoionization, as a function of

59
the photon energy ω, corresponding to a particular bound electron state j
(degenerate with magnetic sublevels mj ) is given by
Z
4π 2 X (−)
σj (ω) = d3 k |hΨj (k) |D · e| Φ0 i|2 δ(ω + E0 − k 2 /2 − ǫj ) . (126)
ωc m
j

Here c ≃ 137 is the speed of light in atomic units and e is the polarization
vector of light. The energy of the system in the initial state E0 + ω and the
final state k 2 /2 + ǫj is conserved. The dipole electromagnetic operator D can
be written in one of the following three forms commonly known as length,
velocity, and acceleration gauges [127]:
 
r ∇ ~1+∇ ~ 2, D =∇˙ 2 r 1 r 2
D = ω(r1 + r2 ), D = ∇ + 3 . (127)
ω r13 r2

The choice of the He atom ground state in Eq. (126) is discussed in detail in
Ref. [128]. The most
P accurate representation is given by a Hylleraas expan-
−zs i j k
sion Φ0 = Ne ijk aijk u s t over the powers of u = |r1 − r2 |, s = r1 + r2 ,
and t = r1 − r2 . With the Hylleraas ground state, all the three gauges of the
dipole operator produce identical cross-section results for the He atom and its
isoelectronic sequence of ions [129]. As an alternative, the ground state can
be represented by a multi-configuration Hartree–Fock (MCHF) expansion.
which is less accurate than the Hylleraas one. However, it can be readily
extended beyond helium and its isoelectronic sequence of ions towards more
complex quasi two-electron targets with a frozen core.
We separate the contribution from the final channels |jk(−) i into single
and double ionization according to the energy of the ǫj which is positive for
the double ionized channels and negative for the singly ionized channels. We
also ensure that the negative-energy state cross sections, which contribute to
the ionization plus excitation cross sections, are multiplied by the projection
of the state onto the true target discrete subspace [43]. This way we have a
relatively clear separation between the discrete and continuous spectrum of
the e-He+ excitation.
The dipole matrix element in Eq. (126) can be used to calculate the
fully resolved triply differential cross section (TDCS) of the DPI [67], which
corresponds to two photoelectrons emerging with energies E1 , E2 within solid
angles Ω1 , Ω2 , respectively. This is achieved by projecting of one, usually
slow, electron onto a positive energy pseudostate of the matching energy

60
ǫj = E2 and by using the partial wave expansion:
(−)
XX
hk2 |jihΨj (k1 )|DM |Φ0 i = (−i)l1 +l2 ei[δl1 (k1 )+δl2 (k2 )]
l1 m1 l2 m2
 
l1 1 l2
×Yl1 m1 (k1 )Yl2 m2 (k2 ) Dl1 l2 (k1 j)hk2 l2 k nj lj i. (128)
m1 M m2

Here, we introduced the reduced (or stripped of angular dependence) dipole

matrix element Dl1 l2 (k1 j) and the overlap integral hk2 l2 k nj lj i between the
positive–energy pseudostate and the Coulomb wave. The former is calculated
by integrating the first-order dipole matrix with the dipole singlet T -matrix
of the e–He+ scattering system. The first-order dipole matrix is obtained as
a transition amplitude of the dipole operator between the correlated ground
state and the final state containing one continuum Coulomb wave and one
discrete pseudostate.
Eq. (128) leads to the following form of TDCS:
2
4π 2 X l1 l2
X
σM = (−i)l1 +l2 Y1M (k̂1 , k̂2 ) e i(δl1 +δl2 ) Dl1 l2 (E1 , E2 ) , (129)
3ωc l l l l
1 2 1 2

l1 l2
where Y1M (k̂1 , k̂2 ) is a bipolar harmonics [130] and

Dl1 l2 (E1 , E2 ) = Dl1 l2 (k1 j)hk2 l2 k nj lj i .

The index M indicates polarization of light and is set to 0 for linearly po-
larized light along the z–direction and to ±1 for circularly polarized light
depending on the helicity.
The TDCS for the linear polarization can be written in a compact form
[131]

4π 2 2
σ0 = (cos θ1 +cos θ2 )Mg (E1 , E2 )+(cos θ1 −cos θ2 )Mu (E1 , E2 ) , (130)
3ωc
where the symmetric and antisymmetric (gerade-g and ungerade-u) DPI am-
plitudes
∞
1 X (−1)l  ′  ±
Mu = g √ Pl+1 (cos θ12 ) ∓ Pl′ (cos θ12 ) Dll+1 (E1 , E2 ) (131)
4π l=0 l + 1

61
depend on the photoelectron energies and their mutual angle cos θ12 = k1 ·
k2 /(k1 k2 ). Following the pioneering work by Schwarzkopf et al. [132], it is
customary to describe the symmetric amplitude by a Gaussian ansatz
"  2 #
π − θ12
Mg (θ12 ) ≈ A exp −2 ln 2 . (132)
∆θ

The full width at half maximum parameter ∆θ depends on the energy of the
photon as prescribed by the Wannier theory [133].

5.1.1. Total integrated cross-sections

The double-to-single photoionization cross-section ratio σ 2+ /σ + in he-
lium has been studied very thoroughly across a wide photon energy range
[134–137]. Our CCC calculation with a 20-parameter Hylleraas ground state
[129], which is illustrated on the right panel of Fig. 17, is essentially gauge
independent and provides a good benchmark for various measurements. This
calculation can be carried out from the threshold, where it reproduces the
Wannier law σ(E) = σ0 E 1.056 , all the way to the limit of infinite photon
energy, indicated by the arrow on the right panel of Fig. 17. This limit is a
static property of the atomic ground state [138]. While the CCC calculation
in the length gauge looses its accuracy in the high photon energy regime, the
two other gauges remain stable and reliable.
Along with the DPI calculation, the CCC model provides the full set
of cross-sections for the ionization accompanied by excitation. These cross-
sections for excitations up to n = 6 ionic states were evaluated for the He
atom and its isoelectronic sequence of ions [129]. The cross section ratios
σn /σ1 for single photoionization of He with excitation to n = 2, 3 and 4 states
are presented in Fig. 18. These data are compared with the experimental
results of Wehlitz et al. [136]. The calculations generally agree very well with
experiment and reach the non-relativistic limit of infinite photon energy.
In a more recent work, Czasch et al. [139] reported partial photoionization
cross sections and angular distributions for double excitation of helium up to
the n = 13 threshold. To reproduce this measurement, the original Laguerre
based CCC method was modified to allow for the use of a box basis, which
discretizes the target spectrum by forcing all target wave functions to vanish
at the box boundary [140].

62
5.1.2. Fully differential cross-sections
The fully resolved TDCS of DPI of helium provide the most stringent test
of theory. The CCC results have been tested in a number of experiments
starting from the benchmark measurement by Bräuning et al. [141] who
established the first good agreement with theory on the absolute scale. These
tests continued in later years across a wide range of photon energies and
polarization state of light which are summarized in Table 1. In all these cases,
agreement between the CCC theory and experiment was very satisfactory.

Table 1: Experimental tests of the

Ref ω, eV Ref ω, eV fully-resolved TDCS of He calcu-
Linear Circular lated by the CCC method at specific
photon energies ω and polarization
[142] 119 [150] 88 state of light.
[143] 139 [151] 127
[144] 159 [152] 179, 529
[145, 146] 119
[147] 529
[148, 149] 179, 529

One of the most stringent tests of the CCC theory was performed in a
measurement of the circular dichroism, i.e. the difference between the TDCS
obtained with left- and right-hand circularly polarized light. Although the
initial measurement by Mergel et al. [153] was found in disagreement with
predictions of the CCC theory [154], subsequent refined measurement by the
same group [155] agreed perfectly well with these predictions as is illustrated
in Fig. 19

5.1.3. Symmetrized amplitudes

The quintessential dynamical information on the DPI process can be con-
densed into the symmetrized amplitudes (131). These amplitudes had been
evaluated by the CCC method both for the case of equal [156] and arbitrary
[157] energy sharing. Even though the Gaussian ansatz (132), as derived
from the Wannier threshold theory, should be strictly valid at small excess
energies, it can be used as a practical tool at a much wider range of photon
energies. The Gaussian width parameters ∆θ deduced from the CCC calcula-
tions at various equal energy sharings are presented in Fig. 20 in comparison
with available experimental data.

63
 Modern experimental techniques allow for direct DPI amplitude measure-
ments without any specific reliance on the Gaussian ansatz [146, 149, 151].
These measurements supported predictions of the CCC theory both for the
gerade and ungerade amplitudes and their relative phase. In a recent devel-
opment [166], it was demonstrated that that an additional information on the
absolute individual phases of the DPI amplitudes can be obtained by a time
delay measurement in a short XUV pulse. Such a measurement, which is
yet to be performed, opens up a possibility of a complete DPI experiment in
which both the magnitudes and phases of the symmetrized DPI amplitudes
can be determined.

5.2. Other targets

The CCC formalism, as outlined in the previous section for helium, can
be readily applied to the isoelectronic sequence of He-like ions [129]. The
only required change is modification of the expansion coefficients aijk in the
Hylleraas ground state and the exponential fall-off parameter in the Laguerre
basis. The same formalism, with the MCHF ground state, can be applied to
the helium atom in its metastable state [167]. More essential modifications
of the formalism is required for other targets.

5.2.1. Quasi two-electron targets

Alkaline-earth metal atoms can be treated by the CCC method in the
so-called frozen-core approximation. In this approximation, all the atomic
electrons, apart from the outermost valence shell, are considered as passive
spectators and form a frozen core. Photoionization of the valence shell is
described by the helium-like formalism with the only modification that both
the target states and the continuum waves entering Eq. (125) are calculated
in the frozen-core Hartree-Fock (FCHF) potential. The atomic ground state
is described by a MCHF expansion which contains two-electron excitations
from the valence shell.
This formalism was applied to Be [168, 169], Mg and Ca [169]. The
double-to-single photoionization cross-sections ratios in Be and Mg are com-
pared with experimental data on the left and right panels of Fig. 21, respec-
tively. Generally, good agreement is achieved with experiment for both tar-
gets. Qualitatively, the double-to-single ratio in alkaline-earth metal atoms
is similar to that of He shown in Fig. 17. These ratios can be scaled to that
of He by measuring the excess energy above the DPI threshold in units of the

64
ionization potential of the corresponding singly charge ion [169]. An expla-
nation of the this scaling property is that the DPI near threshold proceeds
mainly via the electron impact ionization of the singly charged ion. The
cross-section of the former process is a universal function of the reduced ex-
cess energy for all hydrogen-like targets. Qualitative difference of DPI of He
and the alkaline-earth metal atoms can be observed in their angular correla-
tion pattern in two-electron continuum which is described by the Gaussian
function. As compared to He, there is a systematic reduction of the Gaussian
width in the sequence of atoms from Be to Ca [169, 173]. This reduction is
explained by an expansion of the target orbitals in the coordinate space and
corresponding increase of the partial waves of the two-electron continuum
emitted by these orbitals.
Similarly to the valence shell DPI, CCC technique can be applied to the
K-shell DPI of alkaline-earth metal atoms from the threshold to the non-
relativistic limit of infinite photon energy [174]. Theoretical double-to-single
photoionization cross-sections ratios for Mg and Ca are compared favourably
with experimental values derived from high-resolution X-ray spectra follow-
ing the radiative decay of the K-shell double vacancy [175].

5.2.2. Three-electron targets

Recently, the CCC method was expanded to treat the DPI of three-
electron targets like the lithium atom [176]. Application of the CCC method
to DPI of Li is a natural extension of the previous work on DPI of He. This
connection is illustrated in Fig. 22 where we draw schematically the ampli-
tudes of single-photon two-electron ionization of He (left diagram) and Li
(right diagram). Here we use the following graphical symbols. A thin solid
line with an arrow to the right (the direction of time propagation) exhibits
a one-electron state. The dashed line indicates a photon. A shaded oval
stands for the multiple Coulomb interaction summed to infinite order (the T -
matrix). Thick solid lines on the bottom diagram (Li) represent two-electron
states.
Results of CCC calculations of DPI of Li are presented in Fig. 23. Good
agreement with experiment is seen, especially in the near-threshold region
where the calculation clearly adheres to the Wannier threshold law. Except
for the total integrated cross-sections calculations [176], the CCC model was
used to evaluate the differential cross-sections [179] including spin effects in
the final state [180].

65
5.3. Other related processes
The process of electron impact double ionization can be similar to the
DPI. If the incident electron is fast, its interaction with the target can be
treated in the first Born approximation. In this case, the amplitude of the
double ionization process can be presented as a matrix element of the Born
operator between the two-electron ground and final states. Subsequent treat-
ment is identical to that of the DPI with the only difference that instead of
the dipole photoionization operator (127) one has to use the electron im-
pact Born operator. Such a treatment was applied to He at a very high
incident energy of 5.5 keV where the first Born approximation was well jus-
tified [181]. When the CCC calculation was compared with experimental
data, the fully-resolved differential cross-sections were found in good agree-
ment in shape but a considerable rescaling was required to get agreement in
magnitude also. This disagreement in absolute values remains unresolved to
date [182]. Contributions beyond the first Born approximation were identi-
fied comparing experimental data with first order CCC calculations for He
at a smaller incident energy of 2 keV [183] including impulsive regime of
large momentum transfer [184]. Extension of the CCC theory to include the
second-Born corrections was attempted [185]. This technique was also ap-
plied to treat the process of non-sequential two-photon double ionization of
He [186, 187] while the first-Born CCC treatment was given to the process
of electron-impact ionization with simulateneous excitation [122].

6. Concluding remarks
We have presented a surface integral approach to break-up problems that
is valid for both short- and long-ranged potentials. It yields clear defini-
tions of the break-up amplitudes, which have been related to the various
computational methods for calculating such processes. Given how both for-
mal theory and computational techniques have come together, with such a
broad agreement with experiment, it is tempting to conclude that electron-
impact ionization and double photoionization of simple atomic targets are
solved problems. Keeping in mind that single genuine discrepancy between
theory and experiment can bring down the theory we have highlighted some
such cases. We are hopeful that they are due to some as yet undetermined
systematic experimental problem.
Not all questions raised have been answered. The ionization-plus-excitation
problem is currently under intense investigation. Application to more com-

66
plex targets has the added complexity of generating the structure sufficiently
accurately, but within constraints associated with subsequent application to
ionizing collisions.
Nevertheless, we feel that immense progress has occurred in the field
during the last decade, and arguably more than in any other before it. Much
of this was due to the ever growing computational power. As so often happens
in the field of physics, new technologies lead to new levels of understanding. It
has to be acknowledged that the somewhat fortuitous computational progress
was the catalyst for us to revisit the formal theory of breakup with long-range
potentials in order to understand the computational success.

Acknowledgments
We thank Klaus Bartschat, James Colgan, XueGuang Ren, Jochen Röder,
and Mark Baertschy for providing their data in quantitative form. We ac-
knowledge the support of the Australian Research Council, the Australian
National Computational Infrastructure Facility and its Western Australian
node iVEC.

References
[1] T. N. Rescigno, M. Baertschy, W. A. Isaacs, C. W. McCurdy, Colli-
sional breakup in a quantum system of three charged particles, Science
286 (1999) 2474–2479.

[2] M. Baertschy, T. N. Rescigno, C. W. McCurdy, Accurate amplitudes

for electron-impact ionization, Phys. Rev. A 64 (2001) 022709.

[3] P. L. Bartlett, A. T. Stelbovics, I. Bray, Propagating exterior complex

scaling method for electron-hydrogen collisions, J. Phys. B 37 (2004)
L69.

[4] I. Bray, A. T. Stelbovics, Explicit demonstration of the conver-

gence of the close-coupling method for a Coulomb three-body problem,
Phys. Rev. Lett. 69 (1992) 53–56.

[5] I. Bray, Close-coupling approach to coulomb three-body problems,

Phys. Rev. Lett. 89 (2002) 273201.

67
 [6] I. Bray, D. V. Fursa, A. S. Kadyrov, A. T. Stelbovics, Single ionization
of helium by electron impact, Phys. Rev. A 81 (2010) 062704.
[7] K. Bartschat, E. T. Hudson, M. P. Scott, P. G. Burke, V. M. Burke,
Electron-atom scattering at low and intermediate energies using a
pseudo-state/R-matrix basis, J. Phys. B 29 (1996) 115–123.
[8] K. Bartschat, The r-matrix with pseudo-states method: The-
ory and applications to electron scattering and photoionization,
Comp. Phys. Comm. 114 (1998) 168–182.
[9] K. Bartschat, Bsr: B-spline atomic r-matrix codes,
Comp. Phys. Comm. 174 (2006) 273–356.
[10] O. Zatsarinny, K. Bartschat, Nonperturbative treatment of ionization
with excitation of helium by electron impact, Phys. Rev. Lett. 107
(2011) 023203.
[11] J. Colgan, M. S. Pindzola, Double- and triple-differential cross sections
for the low-energy electron-impact ionization of hydrogen, Phys. Rev. A
74 (2006) 012713.
[12] A. S. Kadyrov, A. M. Mukhamedzhanov, A. T. Stelbovics, I. Bray,
Integral representation for the ionization amplitude which is free of
ambiguity and divergence problems, Phys. Rev. Lett. 91 (2003) 253202.
[13] A. S. Kadyrov, A. M. Mukhamedzhanov, A. T. Stelbovics, I. Bray,
Theory of electron-impact ionization of atoms, Phys. Rev. A 70 (2004)
062703.
[14] A. S. Kadyrov, I. Bray, A. M. Mukhamedzhanov, A. T. Stelbovics,
Coulomb breakup problem, Phys. Rev. Lett. 101 (2008) 230405.
[15] A. S. Kadyrov, I. Bray, A. M. Mukhamedzhanov, A. T. Stelbovics,
Surface-integral formulation of scattering theory, Ann. Phys. 324
(2009) 1516–1546.
[16] H. van Haeringen, Charged-particle Interactions, Coulomb Press Ley-
den, Leiden, 1985.
[17] L. D. Faddeev, Scattering theory for a three-particle system, Sov.
Phys. -JETP 12 (1961) 1014–1019.

68
[18] L. D. Faddeev, Mathematical Aspects of the Three-Body Problem in
the Quantum Scattering, Israel Program for Scientific Translations,
Jerusalem, 1965.

[19] E. O. Alt, W. Sandhas, H. Ziegelmann, Coulomb effects in three-body

reactions with two charged particles, Phys. Rev. C 17 (1978) 1981–
2005.

[20] R. K. Peterkop, , Opt. Spectrosc. 13 (1962) 87.

[21] R. G. Newton, Scattering theory of waves and particles, Springer-

Verlag New York, 2nd edition, 1982.

[22] M. Brauner, J. S. Briggs, H. Klar, Triply differential cross sections for

ionisation of hydrogen atoms by electrons and positrons, J. Phys. B 22
(1989) 2265–2287.

[23] S. Jones, D. H. Madison, A. Franz, P. L. Altick, Three-body distorted-

wave Born approximation for electron-atom ionization, Phys. Rev. A
48 (1993) R22–R25.

[24] S. P. Lucey, J. Rasch, C. T. Whelan, On the use of analytic ansatz

three-body wf ..., Proc. R. Soc. A 455 (1999) 349–383.

[25] S. Jones, D. H. Madison, M. Baertschy, Perturbative and nonpertur-

bative calculations of electron-hydrogen ionization, Phys. Rev. A 67
(2003) 012703.

[26] R. K. Peterkop, , Opt. Spektrosk. 13 (1962) 153.

[27] R. K. Peterkop, Theory of Ionization of Atoms by Electron Impact,

Colorado Associated University Press, Boulder, 1977.

[28] G. Arfken, Mathematical Methods for Physicists, Academic Press, New

York, 2nd edition, 1971.

[29] R. K. Peterkop, , Izv. Akad. Nauk Latv. SSR Riga 9 (1960) 79.

[30] C. W. McCurdy, D. A. Horner, T. N. Rescigno, Practical calculation

of amplitudes for electron-impact ionization, Phys. Rev. A 63 (2001)
022711.

69
[31] R. K. Peterkop, Teoriya Ionizatsii Atomov Elektronnym Udarom, Zi-
natne, Riga, 1971.

[32] R. G. Newton, The asymptotic form of three-particle wavefunctions

and cross sections, Ann. Phys. 74 (1972) 324–351.

[33] J. Schwinger, Coulomb green’s function, J. Math. Phys. 5 (1964)

1606–1608.

[34] P. A. M. Dirac, Does renormalization make sense?, in: D. W. Duke,

J. F. Owens (Eds.), AIP Conf. Proc. 74, AIP, New York, 1981, pp.
129–130.

[35] W. Pauli, Nobel Lectures, Physics 1942-1962, Elsevier, Amsterdam,

pp. 27–43.

[36] R. P. Feynman, QED, The Strange Theory of Light and Matter, Prince-
ton University Press, Princeton, pp. 128–129.

[37] V. B. Berestetskii, E. M. Lifshitz, L. P. Pitaevskii, Quantum electro-

dynamics, Pergamon press, 2nd edition, pp. 3–4.

[38] L. H. Ryder, Quantum Field Theory, Cambridge University Press,

Cambridge, 2nd edition, p. 390.

[39] A. Temkin, Nonadiabatic theory of electron-hydrogen scattering, Phys.

Rev. 126 (1962) 130–142.

[40] R. Poet, The exact solution for a simplified model of electron scattering
by hydrogen atoms, J. Phys. B 11 (1978) 3081–3094.

[41] R. Poet, Symmetrically coupled partial differential equations in scat-

tering I. Model electron-hydrogen collisions, J. Phys. B 13 (1980) 2995–
3008.

[42] I. Bray, A. T. Stelbovics, Convergent close-coupling calculations of

electron-hydrogen scattering, Phys. Rev. A 46 (1992) 6995–7011.

[43] I. Bray, A. T. Stelbovics, Calculation of the total ionization

cross section and spin asymmetry in electron-hydrogen scattering,
Phys. Rev. Lett. 70 (1993) 746–749.

70
[44] T. N. Rescigno, M. Baertschy, D. Byrum, C. W. McCurdy, Making
complex scaling work for long-range potentials, Phys. Rev. A 55 (1997)
4253–4262.

[45] C. W. McCurdy, T. N. Rescigno, D. Byrum, Approach to electron-

impact ionization that avoids the three-body coulomb asymptotic form,
Phys. Rev. A 56 (1997) 1958–1969.

[46] G. Gaseano, L. U. Ancarani, D. M. Mitnik, On the applicability of

the exterior complex scaling method for scattering problems including
coulombic potentials, Eur. Phys. J. D (2012) under review.

[47] I. Hornyak, A. T. Kruppa, Two-body coulomb scattering and complex

scaling, Phys. Rev. A 85 (2012) in press.

[48] M. Baertschy, T. N. Rescigno, W. A. Isaacs, C. W. McCurdy, Bench-

mark single-differential ionization cross section results for the s-wave
model of electron-hydrogen scattering, Phys. Rev. A 60 (1999) R13–
R16.

[49] S. Jones, A. T. Stelbovics, Complete numerical solution of Electron-

Hydrogen Model Collision Problem above the Ionization Threshold,
Phys. Rev. Lett. 84 (2000) 1878–1881.

[50] S. Jones, A. T. Stelbovics, Efficient solution of theree-body quantum

collision problems, Phys. Rev. A 66 (2002) 032717.

[51] P. L. Bartlett, I. Bray, S. Jones, A. T. Stelbovics, A. S. Kadyrov,

K. Bartschat, G. ver Steeg, M. P. Scott, P. G. Burke, Unambiguous
ionization amplitudes for electron-hydrogen ionization, Phys. Rev. A
68 (2003) 020702(R).

[52] J. M. Randazzo, F. Buezas, A. L. Frapiccini, F. D. Colavecchia,

G. Gasaneo, Solving three-body-breakup problems with outgoing-flux
asymptotic conditions, Phys. Rev. A 84 (2011) 052715.

[53] D. A. Konovalov, I. Bray, Calculation of electron-impact ionization

using the j-matrix method, Phys. Rev. A 82 (2010) 022708.

[54] J. F. Williams, P. L. Bartlett, I. Bray, A. T. Stelbovics, A. G. Mikosza,

Differential cross sections for excitation to the 3s, 3p and 3d states of

71
 atomic hydrogen by electron impact at energies from 16.5 to 54 eV,
J. Phys. B 39 (2006) 719–728.

[55] P. L. Bartlett, A. T. Stelbovics, Complete direct method for electron-

hydrogen scattering: Application to the collinear and Temkin-Poet
models, Phys. Rev. A 69 (2004) 022703.

[56] G. H. Wannier, Treshold ionization by electron impact ..., Phys. Rev.

90 (1953) 817–825.

[57] P. L. Bartlett, A. T. Stelbovics, Threshold behavior of e-H ionizing

collisions, Phys. Rev. Lett. 93 (2004) 233201.

[58] D. A. Horner, C. W. McCurdy, T. N. Rescigno, Electron-helium

scattering in the S-wave model using exterior complex scaling,
Phys. Rev. A 71 (2005) 012701.

[59] D. A. Horner, C. W. McCurdy, T. N. Rescigno, Electron-impact exci-

tation autoionization of helium in the S-wave limit, Phys. Rev. A 71
(2005) 010701(R).

[60] P. L. Bartlett, A. T. Stelbovics, Electron-helium S-wave model

benchmark calculations. I. Single ionization and single excitation,
Phys. Rev. A 81 (2010) 022715.

[61] P. L. Bartlett, A. T. Stelbovics, Electron-helium S-wave model bench-

mark calculations. II. Double ionization, single ionization with excita-
tion, and double excitation, Phys. Rev. A 81 (2010) 022716.

[62] W. Ihra, M. Draeger, G. Handke, H. Friedrich, Time-dependent

approach to electron scattering and ionization in the s-wave model,
Phys. Rev. A 52 (1995) 3752–3762.

[63] M. S. Pindzola, D. R. Schultz, Time-dependent close-coupling method

for electron-impact ionization of hydrogen, Phys. Rev. A 53 (1996)
1525–1536.

[64] J. Colgan, M. S. Pindzola, F. J. Robicheaux, D. C. Griffin,

M. Baertschy, Time-dependent close-coupling calculations of the triple-
differential cross section for electron-impact ionization of hydrogen,
Phys. Rev. A 65 (2002) 042721.

72
[65] M. S. Pindzola, F. Robicheaux, J. Colgan, Energy and angle differen-
tial cross sections for the electron-impact double ionization of helium,
J. Phys. B 41 (2008) 235202.

[66] J. Colgan, M. Foster, M. S. Pindzola, I. Bray, A. T. Stelbovics, D. V.

Fursa, Triple differential cross sections for the electron-impact ioniza-
tion of helium at 102 ev incident energy, J. Phys. B 42 (2009) 145002
(6pp).

[67] A. S. Kheifets, I. Bray, Convergent close coupling calculations of helium

double photoionization triply differential cross sections, J. Phys. B 31
(1998) L447–L453.

[68] A. S. Kadyrov, I. Bray, Convergence of two-center expansions in

positron-hydrogen collisions, J. Phys. B 33 (2000) L635–L640.

[69] A. S. Kadyrov, I. Bray, A. T. Stelbovics, Near-threshold positron-

impact ionization of atomic hydrogen, Phys. Rev. Lett. 98 (2007)
263202.

[70] I. B. Abdurakhmanov, A. S. Kadyrov, I. Bray, A. T. Stelbovics,

Coupled-channel integral-equation approach to antiproton-hydrogen
collisions, J. Phys. B 44 (2011) 075204.

[71] J. T. Broad, Gauss quadrature generated by diagonalization of H in

finite L2 bases, Phys. Rev. A 18 (1978) 1012–1027.

[72] I. Bray, K. Bartschat, D. V. Fursa, A. T. Stelbovics, Box-based

and Laguerre-based convergent close-coupling calculations of electron-
helium ionisation, J. Phys. B 36 (2003) 3425–3432.

[73] I. Bray, D. V. Fursa, Calculation of ionization within the close-coupling

formalism, Phys. Rev. A 54 (1996) 2991–3004.

[74] E. P. Curran, H. R. J. Walters, Triple differential cross sections for

electron impact ionisation of atomic hydrogen - a coupled psedostate
calculation, J. Phys. B 20 (1987) 337–365.

[75] I. Bray, D. A. Konovalov, I. E. McCarthy, A. T. Stelbovics, Calculation

of triple-differential cross sections in electron scattering on hydrogen,
Phys. Rev. A 50 (1994) R2818–R2821.

73
[76] P. G. Burke, A. Hibbert, W. D. Robb, Electron scattering by complex
atoms, J. Phys. B 4 (1971) 153–161.

[77] P. G. Burke, W. D. Robb, The R-matrix theory of atomic processess,

Adv. Atom. Mol. Phys. 11 (1975) 143–214.

[78] V. M. Burke, C. J. Noble, Farm - a flexible asymptotic r-matrix pack-

age, Comp. Phys. Comm. 85 (1995) 471–500.

[79] M. R. H. Rudge, Theory of the ionization of atoms by electron impact,

Rev. Mod. Phys. 40 (1968) 564–590.

[80] I. Bray, D. V. Fursa, Calculation of singly-differential cross sections of

electron-impact ionization of helium at 100 eV, J. Phys. B 28 (1995)
L435–L441.

[81] I. Bray, Close-coupling theory of ionization: successes and failures,

Phys. Rev. Lett. 78 (1997) 4721–4724.

[82] A. T. Stelbovics, Calculation of ionization within the close-coupling

formalism, Phys. Rev. Lett. 83 (1999) 1570–1573.

[83] K. Bartschat, I. Bray, The s-wave model for electron–hydrogen scat-

tering, Phys. Rev. A 54 (1996) R1002–R1005.

[84] I. Bray, Low energy electron-impact ionization of atomic hydrogen with

equal energy outgoing electrons, J. Phys. B 33 (2000) 581–595.

[85] M. B. Shah, D. S. Elliot, H. B. Gilbody, Pulsed crossed-beam study of

the ionisation of atomic hydrogen , J. Phys. B 20 (1987) 3501–3514.

[86] J. Röder, H. Ehrhardt, C. Pan, A. F. Starace, I. Bray, D. V. Fursa,

Absolute Triply Differential (e,2e) Cross Section Measurements for H
with Comparison to Theory, Phys. Rev. Lett. 79 (1997) 1666–1669.

[87] P. Schlemmer, T. Rösel, K. Jung, H. Ehrhardt, (e,2e) investigation of

atomic hydrogen and helium close to threshold, Phys. Rev. Lett. 63
(1989) 252–254.

[88] J. Röder, M. Baertschy, I. Bray, Ionization of atomic hydrogen by

17.6 ev electrons: experiment and theory, Phys. Rev. A 67 (2003)
010702(R).

74
[89] J. Röder, J. Rasch, K. Jung, C. T. Whelan, H. Ehrhardt, R. J. Allan,
H. R. J. Walters, Coulomb three-body effects in low-energy impact
ionization of h(1s), Phys. Rev. A 53 (1996) 225–233.

[90] C. T. Whelan, R. J. Allan, J. Rasch, H. R. J. Walters, X. Zhang,

J. Röder, K. Jung, H. Ehrhardt, Coulomb three-body effects in
(e,2e) collisions: The ionization of h in coplanar symmetric geometry,
Phys. Rev. A 50 (1994) 4394–4396.

[91] J. Berakdar, J. S. Briggs, I. Bray, D. V. Fursa, Electron angular distri-

butions in low-energy (e,2e) reactions, J. Phys. B 32 (1999) 895–913.

[92] M. Brauner, J. S. Briggs, H. Klar, J. T. Broad, T. Rösel, K. Jung,

H. Ehrhardt, Triply differential cross sections for ionisation of hydrogen
atoms by electrons: the intermediate and threshold energy regions,
J. Phys. B 24 (1991) 657–673.

[93] H. Ehrhardt, K. Jung, G. Knoth, P. Schlemmer, Differential cross

sections of direct single electron impact ionization, Z. Phys. D 1 (1986)
3–32.

[94] J. G. Childers, K. E. James, M. Hughes, I. Bray, M. Baertschy,

M. A. Khakoo, Absolute measurements of electron-hydrogen ioniza-
tion, Phys. Rev. A 68 (2003) 030702(R).

[95] S. Jones, D. H. Madison, M. K. Srivastava, Near-threshold ionization

of hydrogen and helium by electron impact, J. Phys. B 25 (1992)
1899–1914.

[96] I. Bray, Electron-impact ionization of atomic hydrogen at 2 ev above

threshold, J. Phys. B 32 (1999) L119–L126.

[97] I. Bray, On convergence of the close-coupling method for calculating

electron-hydrogen ionisation, J. Phys. B 36 (2002) 2203–2209.

[98] Jochen Röder, Untersuchung der Elektronenstoßionisation, Ph.D. the-

sis, University of Kaiserslautern, Germany, 1996.

[99] D. V. Fursa, I. Bray, Calculation of electron-helium scattering,

Phys. Rev. A 52 (1995) 1279–1298.

75
[100] D. V. Fursa, I. Bray, Topical review: Convergent close-coupling calcu-
lations of electron-helium scattering, J. Phys. B 30 (1997) 757–785.

[101] A. T. Stelbovics, I. Bray, D. V. Fursa, K. Bartschat, Electron-impact

ionization of helium for equal-energy sharing kinematics, Phys. Rev. A
71 (2005) 052716(13).

[102] I. Bray, D. V. Fursa, Benchmark cross sections for electron-impact

total single ionization of helium, J. Phys. B 44 (2011) 061001.

[103] M. S. Pindzola, F. Robicheaux, S. D. Loch, J. C. Berengut, T. Topcu,

J. Colgan, M. Foster, D. C. Griffin, C. P. Ballance, D. R. Schultz,
T. Minami, N. R. Badnell, M. C. Witthoeft, D. R. Plante, D. M. Mit-
nik, J. A. Ludlow, U. Kleiman, The time-dependent close-coupling
method for atomic and molecular collision processes, J. Phys. B 40
(2007) R39–R60.

[104] X. Ren, I. Bray, D. Fursa, J. Colgan, M. Pindzola, T. Pflüger, A. Sen-

ftleben, S. Xu, A. Dorn, J. Ullrich, Electron-impact ionization of he-
lium: A comprehensive experiment benchmarks theory, Phys. Rev. A
83 (2011) 052711.

[105] R. Rejoub, B. G. Lindsay, R. F. Stebbings, Determination of the ab-

solute partial and total cross sections for electron-impact ionization of
the rare gases, Phys. Rev. A 65 (2002) 042713.

[106] A. A. Sorokin, I. L. Beigman, S. V. Bobashev, M. Richter, L. A.

Vainshtein, Total electron-impact ionization cross sections of helium,
J. Phys. B 37 (2004) 3215.

[107] T. Rösel, J. Röder, L. Frost, K. Jung, H. Ehrhardt, Measurement of

absolute, near-threshold triple differential cross sections for electron
impact ionization, J. Phys. B 25 (1992) 3859–3872.

[108] J. Röder, H. Ehrhardt, I. Bray, D. V. Fursa, I. E. McCarthy, Absolute

triple differential cross section for electron-impact ionization of helium
at 40 eV, J. Phys. B 29 (1996) 2103–2114.

[109] S. Rioual, J. Röder, B. Rouvellou, H. Ehrhardt, A. Pochat, I. Bray,

D. V. Fursa, Absolute (e,2e) cross sections for the electron-impact

76
 ionization of helium in energy sharing kinematics at 44.6 eV, J. Phys. B
31 (1998) 3117–3127.

[110] J. Röder, H. Ehrhardt, I. Bray, D. V. Fursa, Absolute double differen-

tial cross sections for electron impact ionization of helium, J. Phys. B
30 (1997) 1309–1322.

[111] J. Röder, H. Ehrhardt, I. Bray, D. V. Fursa, I. E. McCarthy, Absolute

triple differential cross section for electron-impact ionization of helium
at 50 eV, J. Phys. B 29 (1996) L67–L73.

[112] T. Rösel, C. Dupre, J. Röder, A. Duguet, K. Jung, A. Lahmam-

Bennani, H. Ehrhardt, symmetric coplanar e,2e for e-He, J. Phys. B
24 (1991) 3059–3067.

[113] I. Bray, T. Lepage, D. V. Fursa, A. T. Stelbovics, Ionization of helium

by 64.6 ev electrons, J. Phys. B 43 (2010) 074028.

[114] I. Bray, D. V. Fursa, A. S. Kadyrov, A. T. Stelbovics, J. Colgan, M. S.

Pindzola, Close-coupling calculations of 64.6 ev e-he ionization, Journal
of Physics: Conference Series 288 (2011) 012002.

[115] J. Colgan, O. Al-Hagan, D. H. Madison, A. J. Murray, M. S. Pindzola,

Deep interference minima in non-coplanar triple differential cross sec-
tions for the electron-impact ionization of small atoms and molecules,
J. Phys. B 42 (2009) 171001 (6pp).

[116] A. J. Murray, F. H. Read, exploring the helium (e, 2e) differential cross
section at 64.6 ev with symmetric scattering angles but non-symmetric
energies, J. Phys. B 26 (1993) L359–L365.

[117] M. Dürr, C. Dimopoulou, A. Dorn, B. Najjari, I. Bray, D. V. Fursa,

Z. Chen, D. H. Madison, K. Bartschat, J. Ullrich, Single ionization
of helium by 102 ev electron impact: three-dimensional images for
electron emission, J. Phys. B 39 (2006) 4097–4111.

[118] F. Catoire, E. M. Staicu-Casagrande, M. Nekkab, C. D. Cappello,

K. Bartschat, A. Lahmam-Bennani, Investigation of the (e, 2e) single
ionization of he and ar at large energy loss close to minimum momen-
tum transfer, J. Phys. B 39 (2006) 2827–2838.

77
[119] I. Bray, D. V. Fursa, A. T. Stelbovics, Electron-impact ionization of
helium with large energy transfer, Phys. Rev. A 74 (2006) 034702.

[120] L. Avaldi, R. Camilloni, R. Multari, G. Stefani, J. Langlois, O. Robaux,

R. J. Tweed, G. N. Vien, Absolute measurement of ionization-
excitation of helium at intermediate energies and their interpretation,
J. Phys. B 31 (1998) 2981–2997.

[121] P. J. Marchalant, J. Rasch, C. T. Whelan, D. H. Madison, H. R. J. Wal-

ters, First and second born calculations of (e, 2e) excitation-ionization
of helium, J. Phys. B 32 (1999) L705–L710.

[122] A. S. Kheifets, I. Bray, K. Bartschat, Convergent calculations for si-

multaneous electron-impact ionization–excitation of helium, J. Phys. B
32 (1999) L433–L438.

[123] Y. Fang, K. Bartschat, Convergent second-order calculations for simul-

taneous electron-impact ionization-excitation of helium, J. Phys. B 34
(2001) L19–L25.

[124] S. Bellm, J. Lower, K. Bartschat, X. Guan, D. Weflen, M. Foster, A. L.

Harris, D. H. Madison, Ionization and ionization-excitation of helium
to the n = 1 − 4 states of he+ by electron impact, Phys. Rev. A 75
(2007) 042704.

[125] J. Briggs, V. Schmidt, Differential cross sections for photo-double-

ionization of the helium atom, J. Phys. B 33 (2000) R1–R48.

[126] A. S. Kheifets, I. Bray, Calculation of double photoionization of helium

using the convergent close-coupling method, Phys. Rev. A 54 (1996)
R995–R998.

[127] M. Y. Amusia, Atomic photoeffect, Plenum Press, New York, 1990.

[128] A. S. Kheifets, I. Bray, Effect of the ground state correlations on

the helium double photoionization and ionization with excitation,
Phys. Rev. A 57 (1998) 2590–2595.

[129] A. S. Kheifets, I. Bray, Photoionization with excitation and double

photoionization of the helium isoelectronic sequence, Phys. Rev. A 58
(1998) 4501–4511.

78
[130] D. A. Varshalovich, A. N. Moskalev, V. K. Khersonskii, Quantum the-
ory of angular momentum, World Scientific, Singapore, 1988.

[131] A. S. Kheifets, I. Bray, Energy-sharing double photoionization of he-

lium from near-threshold to high energies, Phys. Rev. A 62 (2000)
65402.

[132] O. Schwarzkopf, B. Krässig, J. Elmiger, V. Schmidt, Energy- and

angle-resolved double photoionization in helium, Phys. Rev. Lett. 70
(1993) 3008–3011.

[133] A. Huetz, P. Selles, D. Waymel, J. Mazeau, Wannier theory for double

photoionization of noble gases, J. Phys. B 24 (1991) 1917.

[134] R. Dörner, T. Vogt, V. Mergel, H. Khemliche, S. Kravis, C. L. Cocke,

Ratio of cross sections for double to single ionization of He by 85–400
eV photons, Phys. Rev. Lett. 76 (1996) 2654–2657.

[135] J. C. Levin, G. B. Armen, I. A. Sellin, Photoionization and compton

double ionization of helium from threshold to 20 kev, Phys. Rev. Lett.
76 (1996) 1220–1223.

[136] R. Wehlitz, I. A. Sellin, O. Hemmers, S. B. Whitfield, P. Glans,

H. Wang, D. W. Lindle, B. Langer, N. Berrah, J. Viefhaus, U. Becker,
Photon energy dependence of ionization-excitation in helium at
medium energies, J. Phys. B 30 (1997) L51.

[137] J. A. R. Samson, W. C. Stolte, Z. X. He, J. N. Cutler, R. J. Bartlett,

Double photoionization of helium, Phys. Rev. A 57 (1998) 1906–1911.

[138] T. Aberg, Asymptotic double-photoexcitation cross sections of the

helium atom, Phys. Rev. A 2 (1970) 1726–1729.

[139] A. Czasch, M. Schöffler, M. Hattass, , S. Schössler, T. Jahnke, T. We-

ber, A. Staudte, J. Titze, C. Wimmer, S. Kammer, M. Weckenbrock,
S. Voss, R. E. Grisenti, O. Jagutzki, L. P. Schmidt, H. Schmidt-
Böcking, R. Dörner, J. M. Rost, T. Schneider, C.-N. Liu, I. Bray,
A. Kheifets, K. Bartschat, Partial photoionization cross sections and
angular distributions for double excitation of helium up to the N=13
threshold, Phys. Rev. Lett. 95 (2005) 243003.

79
[140] I. Bray, K. Bartschat, A. T. Stelbovics, Box-based convergent close-
coupling for solving Coulomb few-body problems, Phys. Rev. A 67
(2003) 060704(R).

[141] H. Bräuning, R. Dörner, C. L. Cocke, M. H. Prior, B. Krässig,

A. Kheifets, I. Bray, A. Bräuning-Demian, K. Carnes, S. Dreuil,
V. Mergel, P. Richard, J. Ullrich, H. Schmidt-Böcking, Absolute triple
differential cross sections for photo double ionization of helium — ex-
periment and theory, J. Phys. B 31 (1998) 5149–5160.

[142] S. Cvejanović, J. P. Wightman, T. Reddish, F. Maulbetsch, M. A.

MacDonald, A. S. Kheifets, I. Bray, Photodouble ionisation of helium
at an excess energy of 40 eV, J. Phys. B 33 (2000) 265–283.

[143] C. Dawson, S. Cvejanovic̀, D. P. Seccombe, T. J. Reddish, F. Maul-

betsch, A. Huetz, J. Mazeau, A. S. Kheifets, Helium (γ,2e) triple
differential cross sections at an excess energy of 60 eV, J. Phys. B 34
(2001) L525–L533.

[144] G. Turri, L. Avaldi, P. Bolognesi, R. Camilloni, M. Coreno, J. Berak-

dar, A. S. Kheifets, G. Stefani, Double photoionization of he at 80 ev
excess energy in the equal-energy-sharing condition, Phys. Rev. A 65
(2002) 034702.

[145] P. Bolognesi, R. Camilloni, M. Coreno, G. Turri, J. Berakdar,

A. Kheifets, L. Avaldi, Complementary TDCS for the photo-double
ionisation of He at 40 eV above threshold in unequal energy sharing
conditions, J. Phys. B 34 (2001) 3193–3203.

[146] P. Bolognesi, A. S. Kheifets, I. Bray, L. Malegat, P. Selles, A. K. Kazan-

sky, L. Avaldi, A procedure to extract the complex amplitudes of He
photodouble ionisation from experimental data, J. Phys. B 36 (2003)
L241–L247.

[147] A. Knapp, A. Kheifets, I. Bray, T. Weber, A. L. Landers, S. Schössler,

T. Jahnke, J. Nickles, S. Kammer, O. Jagutzki, L. P. Schmidt,
T. Osipov, J. Rösch, M. H. Prior, H. Schmidt-Böcking, C. L. Cocke,
R. Dörner, Mechanisms of photo double ionization of helium by 530
eV photons, Phys. Rev. Lett. 89 (2002) 033004.

80
[148] A. Knapp, A. Kheifets, I. Bray, T. Weber, A. L. Landers, S. Schössler,
T. Jahnke, J. Nickles, S. Kammer, O. Jagutzki, L. P. Schmidt, T. Os-
ipov, M. H. Prior, H. Schmidt-Böcking, C. L. Cocke, R. Dörner, Photo
double ionization of helium 100 eV and 450 eV above threshold: I.
linearly polarized light, J. Phys. B 38 (2005) 615.

[149] A. Knapp, A. Kheifets, I. Bray, T. Weber, A. L. Landers, S. Schössler,

T. Jahnke, J. Nickles, S. Kammer, O. Jagutzki, L. P. Schmidt, T. Os-
ipov, M. H. Prior, H. Schmidt-Böcking, C. L. Cocke, R. Dörner, Photo
double ionization of helium 100 eV and 450 eV above threshold: III.
Gerade and ungerade amplitudes and their relative phase, J. Phys. B
38 (2005) 645.

[150] A. S. Kheifets, I. Bray, K. Soejima, A. Danjo, K. Okuno, A. Yagishita,

Experimental and theoretical study of linear and circular dichroism in
helium double photoionization, J. Phys. B 32 (1999) L501–L509.

[151] P. Bolognesi, V. Feyer, A. Kheifets, S. Turchini, T. Prosperi, N. Zema,

L. Avaldi, Photodouble ionization of he with circularly polarized
synchrotron radiation: complete experiment and dynamic nodes,
J. Phys. B 41 (2008) 051003.

[152] A. Knapp, A. Kheifets, I. Bray, T. Weber, A. L. Landers, S. Schössler,

T. Jahnke, J. Nickles, S. Kammer, O. Jagutzki, L. P. Schmidt, T. Os-
ipov, M. H. Prior, H. Schmidt-Böcking, C. L. Cocke, R. Dörner, Photo
double ionization of helium 100 eV and 450 eV above threshold: II.
circularly polarized light, J. Phys. B 38 (2005) 635.

[153] V. Mergel, M. Achler, R. Dörner, K. Khayyat, T. Kambara, Y. Awaya,

V. Zoran, B. Nyström, L. Spielberger, J. H. McGuire, J. Feagin, J. Be-
rakdar, Y. Azuma, H. Schmidt-Böcking, Helicity dependence of the
photon-induced three-body coulomb fragmentation of helium investi-
gated by cold target recoil ion momentum spectroscopy, Phys. Rev.
Lett. 80 (1998) 5301–5304.

[154] A. S. Kheifets, I. Bray, Calculation of circular dichroism in helium

double photoionization, Phys. Rev. Lett. 81 (1998) 4588–4591.

[155] M. Achler, V. Mergel, L. Spielberger, R. Dorner, Y. Azuma,

H. Schmidt-Böcking, Photo double ionization of he by circular and

81
 linear polarized single-photon absorption, J. Phys. B 34 (2001) 965–
981.
[156] A. S. Kheifets, I. Bray, Equal energy-sharing double photoionization of
helium from near-threshold to high energies, Phys. Rev. A 62 (2000)
065402.
[157] A. S. Kheifets, I. Bray, Symmetrized amplitudes of the helium atom
double photoionization, Phys. Rev. A 65 (2002) 022708.
[158] O. Schwarzkopf, B. Krassig, J. Elmiger, V. Schmidt, Energy resolved
and angle resolved double photoionization in helium, Phys. Rev. Lett.
70 (1993) 3008–3011.
[159] O. Schwarzkopf, V. Schmidt, Experimental determination of the abso-
lute value of the triple differential cross section for double photoioniza-
tion in helium, J. Phys. B 28 (1995) 2847–2862.
[160] G. Dawber, L. Avaldi, A. G. Mcconkey, H. Rojas, M. A. Macdonald,
G. C. King, Near threshold tdcs for photo-double ionization of helium,
J. Phys. B 28 (1995) L271–L277.
[161] R. Dörner, H. Bräuning, J. M. Feagin, V. Mergel, O. Jagutzki,
L. Spielberger, T. Vogt, H. Khemliche, M. H. Prior, J. Ullrich, C. L.
Cocke, H. Schmidt-Böcking, Photo-double-ionization of He: fully dif-
ferential and absolute electronic and ionic momentum distributions,
Phys. Rev. A 57 (1998) 1074–1090.
[162] L. Malegat, P. Selles, A. Huetz, Double photoionization-1: A new
parametrization of the triple differential cross section from first princi-
ples, J. Phys. B 30 (1997) 251–261.
[163] J. P. Wightman, S. Cvejanović, T. J. Reddish, (γ, 2e) cross section
measurements of D2 and He, J. Phys. B 31 (1998) 1753.
[164] K. Soejima, A. Danjo, K. Okuno, A. Yagishita, Linear and circular
dichroism in the double photoionization of helium, Phys. Rev. Lett. 83
(1999) 1546 – 1549.
[165] L. Malegat, A. Kazansky, P. Selles, Extended Wannier ridge model
versus R-matrix treatment for double photoionization of helium,
J. Phys. B 32 (1999) 4667–4675.

82
[166] A. S. Kheifets, I. A. Ivanov, I. Bray, Timing analysis of two-electron
photoemission, J. Phys. B 44 (2011) 101003.

[167] A. S. Kheifets, A. Ipatov, M. Arifin, I. Bray, Double photoionization

calculations of the helium metastable 2 1,3 s states, Phys. Rev. A 62
(2000) 052724(10).

[168] A. S. Kheifets, I. Bray, Frozen-core model of the double photoionization

of beryllium, Phys. Rev. A 65 (2001) 012710.

[169] A. S. Kheifets, I. Bray, Valence-shell double photoionization of alkaline-

earth-metal atoms, Phys. Rev. A 75 (2007) 042703.

[170] R. Wehlitz, D. Lukic, J. B. Bluett, Single and double photoionization

of beryllium below 40 ev, Phys. Rev. A 71 (2005) 012707.

[171] R. Wehlitz, S. B. Whitfield, Valence double photoionization of beryl-

lium, J. Phys. B 34 (2001) L719–L725.

[172] R. Wehlitz, P. N. Juranić, D. V. Lukić, Double photoionization of

magnesium from threshold to 54 eV photon energy, Phys. Rev. A 78
(2008) 033428.

[173] A. S. Kheifets, I. Bray, Angular correlation in the two-electron contin-

uum, Phys. Rev. A 73 (2006) 020708(R).

[174] A. S. Kheifets, I. Bray, J. Hoszowska, K-shell double photoionization

of be, mg, and ca, Phys. Rev. A 79 (2009) 042504.

[175] J. Hoszowska, A. K. Kheifets, J.-C. Dousse, M. Berset, I. Bray, W. Cao,

K. Fennane, Y. Kayser, M. KavCiC, J. Szlachetko, M. Szlachetko,
Physical mechanisms and scaling laws of k-shell double photoioniza-
tion, Phys. Rev. Lett. 102 (2009) 073006.

[176] A. S. Kheifets, D. V. Fursa, I. Bray, Two-electron photoionization of

ground-state lithium, Phys. Rev. A 80 (2009) 063413.

[177] M.-T. Huang, R. Wehlitz, Y. Azuma, L. Pibida, I. A. Sellin, J. W.

Cooper, M. Koide, H. Ishijima, T. Nagata, Single and double pho-
toionization of lithium, Phys. Rev. A 59 (1999) 3397–3401.

83
[178] R. Wehlitz, J. B. Bluett, S. B. Whitfield, Comparison of the double-
to single-photoionization ratio of li with he, Phys. Rev. A 66 (2002)
012701.

[179] A. S. Kheifets, D. V. Fursa, I. Bray, J. Colgan, M. S. Pindzola, Differ-

ential cross sections of double photoionization of lithium, Phys. Rev.
A 82 (2010) 023403.

[180] A. S. Kheifets, D. V. Fursa, C. W. Hines, I. Bray, J. Colgan, M. S.

Pindzola, Spin effects in double photoionization of lithium, Phys. Rev.
A 81 (2010) 023418.

[181] A. S. Kheifets, I. Bray, A. Lahmam-Bennani, A. Duguet, I. Taouil, A

comparative experimental and theoretical investigation of the electron
impact double ionization of he in kev regime, J. Phys. B 32 (1999)
5047–5065.

[182] A. S. Kheifets, I. Bray, Convergent calculations of double ionization of

helium:
from (γ,2e) to (e,3e) processes, Phys. Rev. A 69 (2004) 050701(R).

[183] A. Dorn, A. Kheifets, C. D. Schröter, B. Najjari, C. Höhr, R. Mosham-

mer, J. Ullrich, Double ionization of helium by electron-impact: Com-
plete pictures of the four-body break up dynamics, Phys. Rev. Lett.
86 (2001) 3755–3758.

[184] A. Dorn, A. Kheifets, C. D. Schröter, B. Najjari, C. Höhr, R. Mosham-

mer, J. Ullrich, Double ionization of helium by electron-impact in the
impulsive regime, Phys. Rev. A 65 (2002) 032709.

[185] A. S. Kheifets, Second-order born model for two-electron atomic ion-

ization by fast charged-particle impact, Phys. Rev. A 69 (2004) 032712.

[186] A. S. Kheifets, I. A. Ivanov, Convergent close-coupling calculations of

two-photon double ionization of helium, J. Phys. B 39 (2006) 1731.

[187] A. S. Kheifets, I. A. Ivanov, I. Bray, Angular anisotropy parameters

and recoil-ion momentum distribution in two-photon double ionization
of helium, Phys. Rev. A 76 (2007) 025402.

84
Figure 16: Coplanar fully differential cross sections n = 1/n = 2 ratios for electron-impact
single ionization of the ground state of helium leaving the residual He+ ion in the n = 1,
or n = 2 states. The two outgoing electrons have E1 = E2 = 44 eV. The experimental
data is due to Bellm et al. [124]. The figure is reproduced with permission of Zatsarinny
and Bartschat [10], who performed the calculations.

85
 4

3
σ2+/σ+ (%)

CCC L
V
2 A
Doerner et al.
1 Wehlitz et al.
Levin et al.
Samson
0
100 150 200 250 300 350 400 100 1000 10000
Photon energy (eV) Photon energy (eV)

Figure 17: The ratio of double to single photoionization cross sections in He. The CCC
calculations are presented in the three gauges of the electro-magnetic interaction with a
20-parameter Hylleraas ground state. Experimental data are from Dörner et al. [134],
Levin et al. [135], Wehlitz et al. [136] and Samson et al. [137]. The arrow on the right
panel visualizes the limit of infinite photon energy.

8.0 L 1.5
V 0.6
σ2/σ1 (%)

σ3/σ1 (%)

σ4/σ1 (%)

7.0 A
Wehlitz et al.
1.0 0.4
6.0

5.0 0.2
0.5
4.0
100 1000 10000 100 1000 10000 100 1000 10000
Photon energy (eV) Photon energy (eV) Photon energy (eV)

Figure 18: The cross sections ratio σn /σ1 for the photoionization of He with simultaneous
excitation to the n = 2, 3 and 4 final ionic state. Experimental data are from Wehlitz
et al. [136]. The asymptotic ratio in the limit of infinite photon energy is indicated by an
arrow.

86
 1
Circular dichroism CD

E1=14-15eV E1=17-18eV E1=19-20eV

0.5

-0.5

-1
60 120 180 240 300 60 120 180 240 300 60 120 180 240 300
Scattering angle θ2 (deg)

Figure 19: Circular dichroism CD=(σ+ −σ− )/(σ+ +σ− ) at excess energy E1 +E2 = 20 eV,
with electron of energy E1 detected at θ1 = 0◦ . Experimental data are from Mergel et al.
[153] (open circles) and Achler et al. [155] (filled circles)

120
Schwarzkopf
Dawber
Width θ12 (deg)

Doerner
100 Malegat
Cvejanovic
Yagishita
Wannier law
80 CCC

60
1 10 100
Excess energy E (eV)

Figure 20: The width parameter ∆θ of the Gaussian function (132). Experimental data
are due to Schwarzkopf et al. [158], Schwarzkopf and Schmidt [159] (open squares), Dawber
et al. [160] (filled squares), Dörner et al. [161] (filled triangles), Malegat et al. [162] (filled
circles), Cvejanović et al. [142, 163] (filled diamonds), Soejima et al. [164] (asterisk). The
CCC results are shown as open circles joint by a solid line. The straight line is the Wannier
exponent by Malegat et al. [165].

87
 1.5

σ++/σ+ (%)
1
σ2+/σ+ (%)

1 CCC (a) 0.5

CCC (b)
Exp. (a) CCC
Exp. (b) Exp
0 0
30 35 40 45 50 25 30 35 40 45
Photon energy ω (eV) Photon energy ω (eV)

Figure 21: Left: Double-to-single photoionization cross-sections ratio in Be (left) and Mg

(right) shown as a function of the photon energy. On the left plot, the CCC results (a) [168]
and (b) [169] are shown by the red filled circles and blue open diamonds. Experimental
data (a) [170] and (b) [171] are displayed by open circles and squares, respectively. On
the right plot, the CCC results [169] are compared with experiment [172].

γ γ
kp k1l 1 1s kp k1 l 1
1s
T 1s ns T
n2 l 2 NS N2 L2
1s ns n’s
2s

Figure 22: (Color online) Graphical representation of the amplitudes of single-photon

two-electron ionization of He (left diagram) and Li (right diagram)

88
 1.5

10
1.0
σ2+/σ+, %

σ2+ (kb)

0.5
CCC 2 CCC
Exp. (a) Exp
Exp. (b) Wannier
0.0 1
80 100 120 140 0.5 1 2 5 10
Photon energy (eV) Excess energy (eV)

Figure 23: (Color online) Left: Double-to-single photoionization cross-sections ratio in Li

is plotted as a function of the photon energy. The CCC calculation [176] is compared with
experiment (a) [177] and (b) [178]. Right: DPI cross-section near threshold is plotted in
log scale as a function of the excess energy. The CCC calculation [176] is compared with
experiment [178] and the Wannier threshold law drawn with a straight line line.

89