Semi-inclusive production of two back-to-back hadron pairs in e+ e− annihilation

revisited
Hrayr H. Matevosyan,1, ∗ Alessandro Bacchetta,2, 3, † Daniël Boer,4, ‡ Aurore Courtoy,5, §
Aram Kotzinian,6, 7, ¶ Marco Radici,3, ∗∗ and Anthony W. Thomas1, ††
1
ARC Centre of Excellence for Particle Physics at the Terascale,
and CSSM, Department of Physics,
The University of Adelaide, Adelaide SA 5005, Australia
http://www.physics.adelaide.edu.au/cssm
2
Dipartimento di Fisica, Università degli Studi di Pavia, via Bassi 6, I-27100 Pavia, Italy, and
3
INFN Sezione di Pavia, via Bassi 6, I-27100 Pavia, Italy
4
Van Swinderen Institute for Particle Physics and Gravity,
University of Groningen, Nijenborgh 4, NL-9747 AG Groningen, The Netherlands
5
Instituto de Fı́sica, Universidad Nacional Autónoma de México,
Apartado Postal 20-364, 01000 Ciudad de México, México.
arXiv:1802.01578v1 [hep-ph] 5 Feb 2018

6
Yerevan Physics Institute, 2 Alikhanyan Brothers St., 375036 Yerevan, Armenia
7
INFN, Sezione di Torino, 10125 Torino, Italy
(Dated: February 7, 2018)
The cross section for back-to-back hadron pair production in e+ e− annihilation provides access to
the dihadron fragmentation functions (DiFF) needed to extract nucleon parton distribution functions
from the semi-inclusive deep inelastic scattering (SIDIS) experiments with two detected final state
hadrons. Particular attention is given to the so-called interference DiFF (IFF), which makes it
possible to extract the transversity parton distribution of the nucleon in the collinear framework.
However, previously unnoticed discrepancies were recently highlighted between the definitions of the
IFFs appearing in the collinear kinematics when reconstructed from DiFFs entering the unintegrated
fully differential cross sections of SIDIS and e+ e− annihilation processes. In this work, to clarify this
problem we re-derive the fully differential cross section for e+ e− annihilation at the leading-twist
approximation. We find a mistake in the definition of the kinematics in the original expression that
systematically affects a subset of terms and that leads to two significant consequences. First, the
discrepancy between the IFF definitions in the cross sections for SIDIS and e+ e− annihilation is
resolved. Second, the previously derived azimuthal asymmetry for accessing the helicity dependent
DiFF G⊥ + −
1 in e e annihilation vanishes, which explains the non-observation of this asymmetry in
the recent experimental searches by the BELLE collaboration. We discuss the recently proposed
alternative option to extract G⊥ 1 .

PACS numbers: 13.60.Hb, 13.60.Le, 13.87.Fh, 12.39.Ki

Keywords: e+ e− to two hadron pair, dihadron fragmentation functions

I. INTRODUCTION proaches have been recently employed in phenomenologi-

The understanding of the complete spin-dependent
structure of the nucleon has been at the forefront of stud-
ies in nuclear physics in recent decades. Particular atten-
tion has been given to studying the so-called transversity
parton distribution function (PDF), which describes the
correlation of the transverse polarization of the nucleon
with the transverse polarization of its constituent par-
tons (see e.g. [1]). The chiral-odd nature of the transver-
sity PDF makes it much harder to measure compared to
the unpolarized and helicity dependent PDFs. Two ap-
∗ ORCID: http://orcid.org/0000-0002-4074-7411
† ORCID: http://orcid.org/0000-0002-8824-8355
‡ ORCID: http://orcid.org/0000-0003-0985-4662
§ ORCID: http://orcid.org/0000-0001-8906-2440
¶ ORCID: http://orcid.org/0000-0001-8326-3284
∗∗
ORCID: http://orcid.org/0000-0002-4542-9797
††
ORCID: http://orcid.org/0000-0003-0026-499X
cal extractions of the transversity [2–5]. The first method
uses the Collins effect [6], that describes the correlation
between the transverse momentum of a produced hadron
with the transverse polarization of an initial quark in the
hadronization process. The convolution upon the trans-
verse momenta of initial and final partons of the transver-
sity and the Collins fragmentation function (FF) can be
measured in a SIDIS process with a single measured fi-
nal state hadron [7], while the convolution of two Collins
FFs are accessible from the semi-inclusive production of
two back-to-back hadrons in e+ e− annihilation [8]. The
second method, based on DiFF, leverages the correlation
between the relative transverse momenta of two produced
hadrons with the transverse polarization of a quark in its
hadronization, which is quantified by the IFF H1^ . Sim-
ilarly to the previous method, here again the SIDIS pro-
cess with two final state hadrons being measured is used
to access a structure function containing the transversity
PDF and an IFF [9–12], while the semi-inclusive produc-
tion of two back-to-back hadron pairs in e+ e− annihila-

tion provides access to IFFs [13–15]. The advantage of
the dihadron method compared to using the Collins effect
is that it is possible to work in the collinear framework
where the corresponding SIDIS structure function factor-
izes in a simple product of the transversity PDF and the
IFF, while for the single hadron case the transversity is
convoluted with the Collins function via an integral in-
volving their transverse momentum dependences. The
same is true for the structure functions containing the
IFF and the Collins FF, respectively, in the e+ e− annihi-
lation cross section. Moreover, in the collinear framework
the same combination of transversity PDF and IFF can
be explored also in proton-proton collisions leading to the
semi-inclusive production of dihadron pairs [16, 17], while
this possibility is in principle precluded for the Collins ef-
fect due to factorization breaking contributions. Finally,
the evolution equations connecting the IFF at different
scales of the various processes have a simple standard
form [18], while the evolution of a transverse-momentum
dependent PDF is more complicated and depends on non
perturbative parameters [19].
A major experimental effort to measure the various az-
imuthal asymmetries involved in extracting the transver-
sity PDF using the dihadron way has been made by sev-
eral collaborations, such as HERMES [20], COMPASS [21, 22],
and BELLE [23, 24]. The IFFs from e+ e− measurements
at BELLE were fitted in Refs. [15, 25]. In turn these
were used in Refs. [4, 25, 26] to successfully extract
the transversity PDF using HERMES and COMPASS data.
Recently, the STAR collaboration released also dihadron
data for azimuthal asymmetries in proton-proton colli-
sions with a transversely polarized proton [27, 28] which
can be included in an attempt of extracting the transver-
sity PDF from a global fit [29].
Recently, systematic model calculations of both FFs
and DiFFs for unpolarized hadrons have been performed
within the extended quark-jet model, which for the
first time provides a self-consistent description for the
hadronization of a quark with an arbitrary polariza-
tion [30–33]. The two DiFF, H1^ and H1⊥ , describing the
correlations between the relative and the total transverse
moment of the hadron pair with the transverse polariza-
tion of the quark, respectively, were studied in Ref. [33].
There, it was observed that the integrated IFF built from
the DiFFs entering the unintegrated SIDIS cross section
is different from the one that is built from the correspond-
ing unintegrated cross section for e+ e− annihilation de-
rived in [13]. In particular, in SIDIS the integrated IFF
contains both the zeroth Fourier cosine moment of the
fully unintegrated H1^ , along with the first Fourier cosine
moment of H1⊥ . This admixture of H1⊥ did not appear
in the original derivation in Ref. [11] but was later in-
cluded in Ref. [34]. On the other hand, the integrated
H1^ in e+ e− annihilation in Ref. [13] contains only the
zeroth Fourier cosine moment of the unintegrated H1^ .
The model estimates of these two definitions of IFFs in
Ref. [33] produced almost a factor of two discrepancy
between them.
2
Another prediction of Ref. [13] concerned a particular
azimuthal modulation that provides access to the first
Fourier cosine moment of the quark helicity dependent
DiFF G⊥ 1 . However, the recent preliminary results from
the BELLE collaboration showed no signal for this modula-
tion within the experimental uncertainties [35, 36]. The
recent COMPASS studies [37] also yielded no significant
signal for SIDIS. Even though the model calculations of
Ref. [32] suggest that the integrated G⊥ 1 appearing in
Ref. [13] is naturally smaller in magnitude than the H1^ ,
this was still a surprise given the precision achieved in
the BELLE analysis.
In this work, we re-derive the unintegrated cross sec-
tion for the semi-inclusive production of two back-to-
back hadron pairs in e+ e− annihilation, first performed
in Ref. [13]. We then re-calculate the azimuthal asymme-
tries used for extracting the IFFs and the helicity depen-
dent DiFF in order to resolve the above discrepancies.
This paper is organized in the following way. In the
next section we briefly review the formalism for DiFF.
In Sec. II we summarize the field-theoretical definitions
of the DiFF. In Sec. III, we describe the kinematics of
two hadron pair production in e+ e− annihilation and re-
derive the corresponding cross section. In Sec. IV, we
re-derive both azimuthal asymmetries involving H1^ and
G⊥1 . We present our conclusions in Sec. V.
II. FIELD-THEORETICAL DEFINITIONS OF
THE DIFFS
The fragmentation of a quark q of an arbitrary polar-
ization s into two unpolarized hadrons h1 , h2 is fully de-
scribed at the leading twist approximation by four DiFFs,
see Refs. [9–11, 13, 34]. The relevant kinematics is de-
scribed by the momentum k and mass m of the quark
q, and the corresponding momenta P1 , P2 and masses
M1 , M2 of the h1 , h2 pair. In the definitions of the DiFFs,
the momenta P1 and P2 of the individual hadrons are re-
placed by their total, P ≡ Ph , and relative, R, momenta
P ≡ Ph = P 1 + P 2 , (1)
1
R = (P1 − P2 ), (2)
2
with Ph2 = Mh2 the squared invariant mass of the pair.
The ẑ axis is defined along the spatial component of
the total momentum Ph and the components of three-
vectors perpendicular to the ẑ direction are denoted by
subscript T , as schematically shown in Fig. 1.
The light-cone momentum fractions of the hadrons are
defined as the ratios of the plus components1 of their
1 The light-cone components of a 4-vector a are defined as a =
(a+ , a− , aT ), where a± = √1 (a0 ± a3 )
2
 3

FIG. 1. The dihadron fragmentation coordinate system,

where the ẑ axis is taken along the total 3-momentum of the
two hadrons, P . The components of 3-momenta perpendicu-
FIG. 2. The kinematics of e+ e− annihilation.
lar to ẑ axis are denoted with a subscript T .

four vectors to the quark momentum, zi = Pi+ /k + . The

following light-cone momentum fractions are used in the D1 (z, ξ, kT2 , RT2 , cos(ϕKR )) (7)
definitions of the DiFFs ∞
1 X cos(n · ϕKR ) [n]
= D1 (z, ξ, |kT |, |RT |),
π n=0 1 + δ 0,n
z = z1 + z2 , (3) and similarly for the other DiFFs.
z1 z2 The invariant mass of the hadron pair Mh is used to
ξ= =1− . (4)
z z replace the magnitude of RT

The two-hadron fragmentation of a quark is described

by a quark-quark correlator [9, 11, 13, 38] RT2 = ξ(1 − ξ)Mh2 − M12 (1 − ξ) − M22 ξ. (8)
These Fourier decompositions will prove valuable when
examining the azimuthal dependence of various structure
∆ij (k; Ph , R) (5)
functions of the e+ e− cross section which we re-derive in
XZ
= d4 ζeik·ζ h0|ψi (ζ)|Ph R, XihPh R, X|ψ̄j (0)|0i, the next section.
X

which, for the case of unpolarized hadron pair and at III. THE e+ e− CROSS SECTION
the leading twist approximation, is parametrized via four
DiFFs In this section we re-derive the e+ e− → h1 h2 +h̄1 h̄2 +X
cross section at the leading twist approximation, follow-
Z ing the framework set out in the original work of Boer
1 et al. [8, 13, 40]. First, we briefly lay out the kinematics
dk − ∆(k,Ph , R)|k+ =P + /z ≡ ∆(z, ξ, kT , RT ) (6)
32z h in the next subsection, followed by the evaluation of the
( cross section itself in the subsequent subsection.
1 1 µνρσ γ µ nν+ kTρ RTσ
= D1 n/ + − G⊥1 γ5
4π 4 Mh2
µ ν µ ν
) A. Kinematics
^ σµν RT n+ ⊥ σµν kT n+
+ H1 + H1 ,
Mh Mh A schematic depiction of the kinematic setup is shown
in Fig. 2. Here, the electron with momentum l annihi-
where D1 is the unpolarized DiFF, G⊥ 1 is the helicity de- lates with a positron of momentum l0 , creating a quark-
pendent DiFF, H1^ is the IFF, and H1⊥ is the analogue antiquark pair. The time like momentum of the interme-
of the Collins function for the dihadron case. The light- diate boson in this hard process is denoted as q = l + l0
like vectors n− and n+ are defined as for any 4-vector and we define q 2 = Q2 . In this work we use Q as the
a, namely a± = a · n∓ , and n+ n− = 1, n2+ = n2− = hard scale and will ignore all the contributions of order
0. All four DiFFs are functions of z, ξ, |kT |, |RT |, and 1/Q. The quark and antiquark hadronize, producing two
kT · RT = |kT ||RT | cos(ϕk − ϕR ), where ϕR and ϕk back-to-back jets. We choose a hadron pair h1 , h2 with
denote the azimuthal angles of the vectors RT and kT . momenta P1 , P2 and masses M1 , M2 from one of the jets.
Thus, the DiFFs only depend on the cosine of the differ- From the other jet, we choose the second hadron pair
ence of the azimuthal angles ϕk − ϕR , that we denote as h̄1 , h̄2 , with momenta P̄1 , P̄2 and masses M̄1 , M̄2 . Here
ϕKR . The DiFFs can be further expanded in an infinite again we define the total and relative transverse momenta
series of Fourier moments with respect to angle ϕKR , as for each pair, as done in Eqs. (1,2), and denote the cor-
done in Ref. [33] (see also Ref. [39] for an alternative ex- responding momenta for the h̄1 , h¯2 pair as P̄h and R̄. In
pansion). It is clear, that all the sine terms vanish, as the ”leading hadron approximation”, where we assume
the DiFFs are even functions of ϕKR . that a significant fraction of the energy in each jet is car-
For D1 we have ried by the two pairs, we can write Ph · P̄h ∼ Q2 . Then
 4

we decompose the momenta Ph , P̄h and q in light-cone The unit vector l̂⊥ can be parametrized using the lepton
coordinates in a frame where PhT = 0 and P̄hT = 0, to plane angle ϕL in the laboratory frame. However, all
identify the corresponding dominant terms the following results are independent of the orientation
of the scattering plane with respect to the laboratory
frame, hence the ϕL dependence will be ignored. Here
Mh2 zh Q µ zh Q µ
Phµ = √ nµ− + √ n+ ≈ √ n+ , (9) we can also define the associated normalized 4-vector
zh Q 2 2 2
z̄h Q M̄h2 z̄h Q µ
P̄hµ = √ nµ− + √ nµ+ ≈ √ n− , (10) ˆl⊥ = l⊥ . (21)
2 z̄h Q 2 2 |l⊥ |
Q Q
q µ = √ nµ− + √ nµ+ + qTµ , (11) Similar to the light-cone frame, we can now define a
2 2 set of orthogonal normalized 4-vectors
where
q
t̂ = , (22)
2Ph · q Q
zh = ≈ z, (12) P̄h
Q2 v̂ = 2 − t̂, (23)
2P̄h · q z̄Q
z̄h = ≈ z̄, (13)
Q2 where the space-like vector v̂ is denoted as ẑ in Refs. [8,
13, 40]. Here we changed the notation to avoid any possi-
and ble confusion with the notation of the ẑ axis. The orthog-
onal projections of the 4-vectors can be again achieved
−qT2 = Q2T  Q2 . (14) using the tensors

We can project the components of 4-vectors transverse µν

g⊥ = g µν − t̂µ t̂ν + v̂ ν v̂ µ , (24)
to n± using the tensors
µν
⊥ = − µνρσ
t̂ρ v̂σ , (25)

gTµν = g µν − nµ+ nν− − nν+ nµ− , (15) The two perpendicular projection tensors can be re-
lated
µν
T =
µνρσ
n+ρ n−σ , (16)

where g µν is the metric tensor. µν nµ+ qTν + nν+ qTµ

The coordinate system in Fig. 2 is defined by taking the g⊥ = gTµν − . (26)
Q
ẑ axis opposite to the 3-momentum P̄h , while the com-
ponents of the vectors perpendicular to ẑ are denoted In this work we neglect all terms of order QT /Q, Mh /Q,
with a subscript ⊥ in a frame where q⊥ = 0. It can be M̄h /Q. Thus we also neglect the differences between the
easily shown, that Ph⊥ = −zqT , up to negligible correc- T and ⊥ components of vectors.
tion of order Q2T /Q2  1. We can then define the two
orthogonal unit vectors in ⊥ direction
B. Cross section

Ph⊥ qT
ĥ = =− , (17) The cross section for this process is given by the con-
|Ph⊥ | |qT |
volution of leptonic and hadronic tensors
ĝ i = ij j
T ĥ = 
0ij3 j
ĥ , (18)

where the following convention is used 0123 = +1. 2P10 2P20 2P̄10 2P̄20 dσ α2 µν
= Lµν W(4h) , (27)
To keep consistency, we will define all the azimuthal d3 P1 d3 P2 d3 P̄1 d3 P̄2 Q6
angles with respect to the lepton frame. Then, we can where
parametrize these two vectors using the azimuthal angle
φ1 of ĥ "
2 µν
Lµν = Q − 2A(y)g⊥ (28)
ĥ = (cos(φ1 ), sin(φ1 )), (19)
ĝ = (sin(φ1 ), − cos(φ1 )), (20)

µ ˆν 1 µν 
+ 4B(y)v̂ µ v̂ ν − 4B(y) ˆl⊥ l⊥ + g⊥
2
#
so that the azimuthal angle of ĝ is simply φg = 3/2π+φ1 .  
The lepton plane in Fig. 2 is spanned by the ẑ axis and − 2C(y)B 1/2 (y) v̂ µ ˆlν + v̂ ν ˆlµ ,
⊥ ⊥
the transverse component l⊥ of the electron momentum l.
 5

and with

Ph · l l− 1 + cos θ2
y= ≈ − = . (32)
1 Ph · q q 2
A(y) = − y + y2 , (29)
2
The last equality holds in the center-of-mass frame,
B(y) = y(1 − y), (30)
where θ2 is the angle between the 3-momentum of the
C(y) = 1 − 2y, (31) electron l and the ẑ axis.

The hadronic tensor is defined as

d3 PX
Z
µν 1 X 4
W(4h) (q; Ph , R, P̄h , R̄) =
(2π)10 0 (2π) δ(q − PX − Ph − P̄h )
(2π)3 2PX
(33)
X
× h0|J µ (0)|PX ; Ph , R, P̄h , R̄ihPX ; Ph , R, P̄h , R̄|J ν (0)|0i.

¯
Using the parton picture, we can decompose the hadronic tensor in terms of the quark-quark correlators ∆ and ∆
for the production of the two hadron pairs in the fragmentation of the quark and the antiquark

Z " #
µν
X
W(4h) ≈ 3(32z)(32z̄) e2a 2 2 2 ¯
d kT d k̄T δ (qT − kT − k̄T )Tr ∆(z̄, ¯ k̄T , R̄T )γ ∆(z, ξ, kT , RT )γ
ξ, µ µ
, (34)
a,ā

where a denotes the flavor of the fragmenting quark and the pre factor is the number of active colors NC = 3.

Following the transformation of the phase space factor detailed in Ref. [13], the cross section expression can be
written as

dσ α2
= z z̄Lµν W µν . (35)
d2 qT dz dξ dϕR dMh2 dz̄ dξ¯ dϕR̄ dM̄h2 dy dϕL 128Q4

Up until this point we have followed the same formalism and definitions as in Ref. [13]. The next step is to evaluate
the trace in Eq. (34) and contract the resulting expression for the hadronic tensor with the leptonic tensor in Eq. (28).
The resulting expression follows
 6

 
dσ e+ e− → (h1 h2 )(h̄1 h̄2 )X 3α2 2 2 X 2
2 ¯ 2
= z z̄ ea (36)
2
d qT dzdξdϕR dMh dz̄dξdϕR̄ dM̄h dy πQ2 a,ā
( " # " #
a ā |RT | |R̄T | ^a ^ā
× A(y)F D1 D̄1 + cos(ϕR + ϕR̄ )B(y) F H1 H̄1
Mh M̄h
" #
B(y)  
⊥a ⊥ā
+ cos(2φ1 ) F 2(ĥ · kT )(ĥ · k̄T ) − (kT · k̄T ) H1 H̄1
Mh M̄h
" #
B(y)  
⊥a ⊥ā
+ sin(2φ1 ) F (ĝ · kT )(ĥ · k̄T ) + (ĥ · kT )(ĝ · k̄T ) H1 H̄1
Mh M̄h
" # " #
B(y)|RT | ^a ⊥ā B(y)|RT | ^a ⊥ā
+ cos(φ1 + ϕR ) F (ĥ · k̄T )H1 H̄1 + sin(φ1 + ϕR ) F (ĝ · k̄T )H1 H̄1
Mh M̄h Mh M̄h
" # " #
B(y)|R̄T | ⊥a ^ā B(y)|R̄T | ⊥a ^ā
+ cos(φ1 + ϕR̄ ) F (ĥ · kT )H1 H̄1 + sin(φ1 + ϕR̄ ) F (ĝ · kT )H1 H̄1
Mh M̄h Mh M̄h
" #
|RT | |R̄T |  
⊥a ⊥ā
− A(y) 2 sin(φ1 − ϕR ) sin(φ1 − ϕR̄ )F (ĥ · kT )(ĥ · k̄T ) G1 Ḡ1
Mh M̄h2
" #
 
− sin(φ1 − ϕR ) cos(φ1 − ϕR̄ )F (ĥ · kT )(ĝ · k̄T ) G⊥a 1 Ḡ1
⊥ā

" #
 
− cos(φ1 − ϕR ) sin(φ1 − ϕR̄ )F (ĝ · kT )(ĥ · k̄T ) G⊥a ⊥ā
1 Ḡ1

" #!)
 
+ cos(φ1 − ϕR ) cos(φ1 − ϕR̄ )F (ĝ · kT )(ĝ · k̄T ) G⊥a ⊥ā
1 Ḡ1 ,

where the convolution F is defined as

Z
F[wDa D̄ā ] = ¯ k̄2 , R̄2 , k̄T · R̄T ).
d2 kT d2 k̄T δ 2 (kT + k̄T − qT )w(kT , k̄T , RT , R̄T )Da (z, ξ, kT2 , RT2 , kT · RT )Dā (z̄, ξ, T T

(37)

There are several important differences between the expression in Eq. (36) and the original expression in Eq. (19) of
Ref. [13], apart from the different mass normalization. First, the terms multiplying ĝ are multiplied by a factor of −1
in our expression. Second, the factor A(y) in front of the G⊥a ⊥ā
1 Ḡ1 terms is also multiplied by a factor of −1. Lastly,
the dependence on angle ϕL vanishes altogether, as in this work all the azimuthal angles are defined with respect to
the lepton plane.
These differences allow us to rewrite the cross section in a much more compact form

 
dσ e+ e− → (h1 h2 )(h̄1 h̄2 )X
( " #
3α2 2 2 X 2 a ā
¯ R̄ dM̄ 2 dy = πQ2 z z̄
d2 qT dzdξdϕR dMh2 dz̄dξdϕ
ea A(y)F D1 D̄1 (38)
h a,ā
" # " #
|kT | |k̄T | ⊥a ⊥ā |RT | |R̄T | ^a ^ā
+ B(y)F cos(ϕk + ϕk̄ )H1 H̄1 + B(y)F cos(ϕR + ϕR̄ )H1 H̄1
Mh M̄h Mh M̄h
" # " #
|kT | |R̄T | ⊥a ^ā |RT | |k̄T | ^a ⊥ā
+ B(y)F cos(ϕk + ϕR̄ )H1 H̄1 + B(y)F cos(ϕR + ϕk̄ )H1 H̄1
Mh M̄h Mh M̄h
" #)
|RT | |kT | |R̄T | |k̄T | ⊥a ⊥ā
− A(y)F sin(ϕk − ϕR ) sin(ϕk̄ − ϕR̄ )G1 Ḡ1 .
Mh2 M̄h2
 7

We obtain the cross section in collinear kinematics by integrating upon d2 qT . This integration trivially breaks up
the convolution between kT and k̄T in Eq. (37). In the last line, we have the product of two terms of the following
form

Z
dϕk sin(ϕKR )G⊥a
1 (z, ξ, |kT |, |RT |, cos(ϕKR )) = 0, (39)

that trivially vanishes by changing the integration variable ϕk → ϕKR . By replacing in Eq. (38) the remaining DiFFs
with their Fourier cosine decompositions in Eq. (7), we have

 
dσ e+ e− → (h1 h2 )(h̄1 h̄2 )X 3α2 2 2
Z Z X cos(nϕKR ) cos(mϕ ) X
2 2 K̄ R̄
¯ R̄ dM̄ 2 dy = z z̄ d k T d k̄T e2a (40)
dzdξdϕR dMh2 dz̄dξdϕ h πQ2 n,m
π(1 + δ 0,n ) π(1 + δ 0,m ) a,ā
(
a,[n] ā,[m] B(y) ⊥a,[n] ⊥ā,[m] ^a,[n] ^ā,[m]
× A(y)D1 D̄1 + cos(ϕk + ϕk̄ )|kT |H1 |k̄T |H̄1 + cos(ϕR + ϕR̄ )|RT |H1 |R̄T |H̄1
Mh M̄h
!)
⊥a,[n] ^ā,[m] ^a,[n] ⊥ā,[m]
+ cos(ϕk + ϕR̄ )|kT |H1 |R̄T |H̄1 + cos(ϕk̄ + ϕR )|RT |H1 |k̄T |H̄1
(
3α2 2 2 dkT2 dk̄T2 X 2
Z Z
a,[0] ā,[0] B(y) 
⊥a,[1] ⊥ā,[1]
= z z̄ ea A(y)D1 D̄1 + cos(ϕR + ϕR̄ ) |kT |H1 |k̄T |H̄1
πQ2 2 2 a,ā Mh M̄h
)

^a,[0] ^ā,[0] ⊥a,[1] ^ā,[0] ^a,[0] ⊥ā,[1]
+ |RT |H1 |R̄T |H̄1 + |kT |H1 |R̄T |H̄1 + |RT |H1 |k̄T |H̄1
( )
3α2 X 2 a 2 ā ¯ 2 |RT | |R̄T | ^a 2 ^ā ¯ 2
= e A(y)D1 (z, ξ, Mh )D̄1 (z̄, ξ, M̄h ) + B(y) cos(ϕR + ϕR̄ ) H (z, ξ, Mh )H̄1 (z̄, ξ, M̄h ) ,
4π 3 Q2 a,ā a Mh M̄h 1

where

dkT2 [0]
Z Z
[0]
D1 (z, ξ, Mh2 ) ≡ z 2 d2 kT D1 (z, ξ, |kT |, |RT |) = 2πz 2 D1 (z, ξ, |kT |, |RT |), (41)
2
|kT | ⊥,[1]
Z h i
^,[0]
H1^ (z, ξ, Mh2 ) ≡ z 2 d2 kT H1 (z, ξ, |kT |, |RT |) + H1 (z, ξ, |kT |, |RT |) (42)
|RT |
2 h
|kT | ⊥,[1]
Z
dkT ^,[0]
i
= 2πz 2 H1 (z, ξ, |kT |, |RT |) + H1 (z, ξ, |kT |, |RT |) ,
2 |RT |

and similarly for the barred functions.

Following Ref. [12], we can expand the above DiFF in the relative partial waves of the hadron pair system. In the
center-of-mass (c.m.) frame of the pair, we can change the ξ dependence to ζ = 2ξ − 1 = a + b cos θ, where a, b are
functions only of Mh2 and θ is the angle between the direction of the back-to-back emission in the c.m. frame and the
direction of Ph in the target rest frame. The Jacobian of the transformation is dξ = |R|/Mh d cos θ. Then, we have

|R|
2 D1 (z, ξ, Mh2 ) = D1 (z, cos θ, Mh2 ) = D1,OO (z, Mh2 ) + cos θD1,OL (z, Mh2 ) + . . . (43)
Mh
|R| ^
2 H (z, ξ, Mh2 ) = H1^ (z, cos θ, Mh2 ) = H1,OT
^
(z, Mh2 ) + cos θH1,LT
^
(z, Mh2 ) + . . . (44)
Mh 1

If we insert these expansions in Eq. (40) retaining only the first non-vanishing term after integrating in d cos θ (d cos θ̄),
and we further change the y variable as in Eq. (32), then the collinear cross section can be written as
 8

 
dσ e+ e− → (h1 h2 )(h̄1 h̄2 )X 1 3α2 1 + cos2 θ2 X 2 a
= ea D1,OO (z, Mh2 )D̄1,OO
ā
(z̄, M̄h2 ) (45)
dzd cos θdϕR dMh2 dz̄d cos θ̄dϕR̄ dM̄h2 d cos θ2 4π 2 8πQ2 4 a,ā
( )
2 ^a 2 ^ā 2
2
P
sin θ2 |R| |R̄| a,ā ea H1,OT (z, Mh )H̄1,OT (z̄, M̄h )
× 1 + cos(ϕR + ϕR̄ ) sin θ sin θ̄ a
1 + cos2 θ2 (z, Mh2 )D̄1,OO
ā (z̄, M̄h2 )
P
Mh M̄h a,ā e2a D1,OO
1 h i
= 2
dσ0 1 + cos(ϕR + ϕR̄ )A(cos θ2 , cos θ, cos θ̄, z, Mh2 , z̄, M̄h2 ) ,
4π
where RT = R sin θ (and similarly for R̄T ), dσ0 is the unpolarized cross section, and A is the so-called Artru–Collins
asymmetry.
The above expression is identical (up to a numerical factor) to the one used in Ref. [15] to extract the IFF from the
BELLE experimental data for the Artru–Collins asymmetry [23]. The same IFF occurs also in the SIDIS cross section
for the semi–inclusive production of hadron pairs off transversely polarized targets [12], and it is used to extract the
transversity distribution from a suitable single-spin asymmetry [4, 25, 26]. Without expanding the DiFF in relative
partial waves and by directly computing the cos(ϕR + ϕR̄ ) moment of the cross section in Eq. (40), the resulting
Artru–Collins asymmetry is also formally identical to that in Eq. (23) of Ref. [13] (see next section). The crucial
difference is in the definition of Eq. (42), namely in how the integrated IFF entering the asymmetry is built in terms
of unintegrated DiFF. Starting from the correct cross section of Eq. (36), the expression in Eq. (42) (multiplied by
|RT |) is now consistent with the definition of IFF entering the azimuthal asymmetry in the SIDIS cross section [33]
(see also Ref. [34]). The same consistency could not be achieved from the cross section in Eq. (19) of Ref. [13]. Thus,
the discrepancy is indeed resolved.

IV. THE AZIMUTHAL ASYMMETRIES

In this section, we will review and discuss the azimuthal asymmetries that allow to extract the IFF and the helicity
dependent DiFF from the cross section listed in Eq. (38). For this purpose, we define the average of an arbitrary
function I as

 
Z Z Z Z Z dσ e+ e− → (h1 h2 )(h̄1 h̄2 )X
hIi ≡ dξ dξ¯ dϕR dϕR̄ d 2 qT I 2 ¯ R̄ dM̄ 2 dy . (46)
d qT dzdξdϕR dM 2 dz̄dξdϕ
h h

We first calculate the integral of the unweighted cross section, that appears as denominator in all of the azimuthal
asymmetries. Following the same steps leading to Eq. (40), we have

 
Z Z Z Z Z dσ e+ e− → (h1 h2 )(h̄1 h̄2 )X
h1i = dξ dξ¯ dϕR dϕR̄ d 2 qT ¯ R̄ dM̄ 2 dy (47)
d2 qT dzdξdϕR dM 2 dz̄dξdϕ
h h
3α2 X
= A(y) e2a D1a (z, Mh2 )D̄1ā (z̄, M̄h2 ),
πQ2 a,ā

where

Z
D1a (z, Mh2 ) = dξ D1a (z, ξ, Mh2 ), (48)

and D1 (z, ξ, Mh2 ) is given in Eq. (41) (and similarly for D̄1ā ).

A. Artru-Collins asymmetry

In Ref. [13], the Artru-Collins asymmetry is defined as

 9

hcos(ϕR + ϕR̄ )i
A(y, z, z̄, Mh2 , M̄h2 ) = . (49)
h1i

Following the same steps leading to Eq. (40), we have

3α2 X
hcos(ϕR + ϕR̄ )i = B(y) e2a H1^a (z, Mh2 )H̄1^ā (z̄, M̄h2 ), (50)
2πQ2 a,ā

where

|RT | ^
Z
H1^ (z, Mh2 ) = dξ H (z, ξ, Mh2 ), (51)
Mh 1

with H1^ (z, ξ, Mh2 ) given in Eq. (42) (and similarly for H̄1^ā ).
Finally, the Artru–Collins asymmetry results

" #" #−1

1 X X
A(y, z, z̄, Mh2 , M̄h2 ) = B(y) e2a H1^a (z, Mh2 )H̄1^ā (z̄, M̄h2 ) A(y) e2a D1a (z, Mh2 )D̄1ā (z̄, M̄h2 ) , (52)
2 a,ā a,ā

which is identical to Eq. (23) of Ref. [13], but now H1^a (z, Mh2 ) is given by Eq. (51) consistently with the definition
entering the azimuthal asymmetry in the SIDIS cross section [33] (and similarly for H̄1^ā (z̄, M̄h2 )).

B. The asymmetry for the helicity dependent DiFF

Another important consequence of the new expression for the cross section in Eq. (38) is that the so-called longitu-
dinal jet handedness azimuthal asymmetry, suggested in Ref. [13] to address the helicity dependent DiFF, identically
vanishes. This asymmetry is defined as

hcos(2(ϕR − ϕR̄ ))i

A⇒ (y, z, z̄, Mh2 , M̄h2 ) = . (53)
h1i

The contributions to hcos(2(ϕR − ϕR̄ )i from terms in Eq. (38) involving B(y) vanish, which is easy to check using
similar steps to those used in the derivations of Eq. (40), where we quickly end up with an expression multiplied by

Z Z
dϕR dϕR̄ cos(2(ϕR − ϕR̄ )) cos(ϕR + ϕR̄ ) = 0. (54)

The only remaining contribution is by the last term in Eq. (38), which we can again transform to a much simpler
from by redefining ϕk → ϕKR , ϕk̄ → ϕK̄ R̄ after integrating upon dqT :

3α2 X Z Z Z Z
h cos(2(ϕR − ϕR̄ ))i = − 2
A(y) e2a dξ dξ¯ dϕR dϕR̄ cos(2(ϕR − ϕR̄ )) (55)
πQ a,ā

|kT ||RT | ⊥a   |k̄ ||R̄ |

Z Z  
T T ⊥ā ¯
× z 2 z̄ 2 dkT dk̄T sin(ϕk ) sin(ϕk̄ ) G1 z, ξ, |kT |, |RT |, cos(ϕk ) Ḡ1 z̄, ξ, | k̄T |, |R̄T |, cos(ϕ k̄ )
Mh2 M̄h2
= 0.

Thus, the asymmetry of Eq. (53) identically vanishes. In fact, any moment of the cross section that depends only on
angles ϕR and ϕR̄ would get no contribution from the terms involving G⊥
1 , as can readily be seen from the derivation
in Eq. (55) since the integration upon dqT already yields a zero.
 10

It is interesting to investigate if there is a specific moment that allows to single out the helicity dependent DiFF
G⊥
1 . If we include in the weight information on |qT |, following the same steps as before for example we get

(
3α2 X a,[1],(1/2) ā,[1],(1/2)
hqT2 cos(ϕR − ϕR̄ )i = A(y) 2
ea Mh M̄h 2D1 (z, Mh2 )D̄1 (z̄, M̄h2 ) (56)
πQ2 a,ā
)
  
⊥a,[0],(1) ⊥a,[2],(1) ⊥ā,[0],(1) ⊥ā,[2],(1)
− G1 (z, Mh2 ) − G1 (z, Mh2 ) Ḡ1 (z̄, M̄h2 ) − Ḡ1 (z̄, M̄h2 )
( )
3α2 X a,[1],(1/2) ā,[1],(1/2)
≡ A(y) e2a Mh M̄h 2D1 (z, Mh2 )D̄1 (z̄, M̄h2 ) − G⊥a 2 ⊥ā 2
1 (z, Mh )Ḡ1 (z̄, M̄h ) ,
πQ2 a,ā

where

p Z
kT2
Z 
[n],(p) [n]
D1 (z, Mh2 ) ≡ z 2 d2 kT dξ D1 (z, ξ, |kT |, |RT |), (57)
2Mh2
p Z
kT2

|RT | ⊥,[n]
Z
⊥,[n],(p)
G1 (z, Mh2 ) ≡z 2 2
d kT dξ G (z, ξ, |kT |, |RT |), (58)
2Mh2 Mh 1
⊥,[0],(1) ⊥,[2],(1)
G⊥ 2
1 (z, Mh ) ≡ G1 (z, Mh2 ) − G1 (z, Mh2 ), (59)

are kT2 −moments of order p of the Fourier cosine moments of order n of the involved DiFF (and similarly for the
barred functions). Note, that this definition of G⊥ 2
1 (z, Mh ) is different than that in Ref [13]. Therefore, weighing
2
the cross section with a function of ϕR , ϕR̄ and qT is not enough to isolate its contribution coming from the helicity
dependent DiFF.
Such new weight has been recently proposed in Ref. [41], that also involves the azimuthal angle ϕq = ϕ1 + π of qT
to exactly cancel out the contributions from the unpolarized term in the cross section:

D  E
qT2 3 sin(ϕq − ϕR ) sin(ϕq − ϕR̄ ) + cos(ϕq − ϕR ) cos(ϕq − ϕR̄ ) (60)
D  E
= qT2 2 cos(ϕR − ϕR̄ ) − cos(2ϕ1 − ϕR − ϕR̄ )
12α2 X
= 2
A(y) e2a Mh M̄h G⊥a 2 ⊥ā 2
1 (z, Mh ) Ḡ1 (z̄, M̄h ),
πQ a,ā

where G⊥ 2 ⊥ 2
1 (z, Mh ) is defined in Eq. (59) (and similarly for Ḡ1 (z̄, M̄h ) ).
Finally, it is worth noticing that since hqT2 cos(ϕR − ϕR̄ )i = 6 0 and hcos(ϕR − ϕR̄ )i = 0, the latter moment can
contain terms that survive the integration upon ϕq but vanish because of the integration upon the modulus |qT |. If
we perform all the integrations indicated in Eq. (46) except for the one upon d|qT |, the only surviving contribution
is (see Appendix A)

3α2 X Z
hcos(ϕR − ϕR̄ )i(qT2 ) = 2
A(y) e2a dϕq F1a (qT2 , z, z̄, RT2 , R̄T2 ) (61)
πQ a,ā

3α2 X
= 2
A(y) e2a 2π F1a 6= 0
πQ a,ā

where

Z Z
F1a (qT2 , z, z̄, RT2 , R̄T2 ) = d2 kT d2 k̄T δ(kT + k̄T − qT ) cos(ϕk − ϕk̄ )
(Z
a,[1]
× dξD1 (z, ξ, |kT |, |RT |)
|kT ||RT |  ⊥a,[0]
Z
1 ⊥a,[2]
− dξ 2 G1
4 Mh
| ||
Z
k̄ T R̄T
× dξ¯
M̄h2
If hcos(ϕR − ϕR̄ )i = 0 vanishes because of the integration upon the modulus |qT |, it means that this moment,
when considered as a function of qT2 , must have a node. Indeed, some preliminary measurements from the BELLE
collaboration indicate a non vanishing hcos(ϕR − ϕR̄ )i which could be due to the limited coverage in qT2 [42]. However,
it is not evident which combination of moments of DiFF in Eq. (62) is responsible for a node in Eq. (61). In principle,
both terms could contribute in changing the sign of hcos(ϕR − ϕR̄ )i because the Fourier cosine moment D1 is not
necessarily a positive definite function.
V. CONCLUSIONS
The DiFF provide a very rich source of information
concerning the hadronization process. Moreover, in re-
cent years they have been used to explore the structure
of the nucleon using two-hadron semi-inclusive electro-
production. The information about the DiFF extracted
from the two back-to-back hadron pair semi-inclusive
production in e+ e− annihilation plays an absolutely vi-
tal role in these studies. The fully unintegrated cross
section for this process and the relevant azimuthal asym-
metries for accessing the different DiFF were first derived
in Ref. [13].
We recently observed in Ref. [33] that the integrated
IFF built from the DiFF entering the unintegrated SIDIS
cross section is apparently different from the one that is
built from the corresponding unintegrated cross section
for e+ e− annihilation obtained in Ref. [13]. In this work
we re-derived these quantities following the same kine-
matic setup of Ref. [13]. In Sec. III B, we found a mis-
take in the definition of the kinematics that impacts a
subset of terms in the cross section having significant im-
plications for the relevant asymmetries. The most impor-
tant result derived in Sec. IV A is that with the corrected
cross section the apparent discrepancy between the def-
initions of the integrated IFF in terms of unintegrated
DiFF occurring in the SIDIS and e+ e− cross sections is
resolved. Although the procedure used in the extraction
of the transversity PDF using the dihadron method in
Refs. [4, 25, 26] is formally correct, it is nevertheless im-
portant to have a consistent underlying formalism, which
has been established here.
The second important result, derived in Sec. IV B,
is that that azimuthal asymmetry previously proposed
for accessing the helicity dependent DiFF G⊥ 1 actually
vanishes. The reason is the complete decoupling of the
quark and antiquark transverse momenta in these asym-
metries, as a consequence of which the modulations of
11
(62)
Z
ā,[1]
dξ¯D̄1 ¯ |k̄T |, |R̄T |)
(z̄, ξ,

(z, ξ, |kT |, |RT |) − G1 (z, ξ, |kT |, |RT |)
)
|

⊥ā,[0] ⊥ā,[2]

Ḡ1 ¯ |k̄T |, |R̄T |) − Ḡ
(z̄, ξ, 1
¯ |k̄T |, |R̄T |) .
(z̄, ξ,
[1]
their respective hadron productions are lost. This nat-
urally explains the absence of the corresponding signal
in the recent analysis at BELLE [35, 36]. Further, we
discussed the azimuthal asymmetry recently proposed in
Ref. [41] that allows to access G⊥ 1 . We have also ana-
lyzed another azimuthal asymmetry based on the relative
azimuthal orientation of the planes containing the two
back-to-back hadron pair momenta. Interestingly, this
asymmetry vanishes independently of the various angu-
lar integrations, because it displays a node as a function
of the size of the imbalance between the transverse mo-
menta of the two back-to-back jets. As a consequence,
incomplete integration on the imbalance size would gen-
erate a non-vanishing result, as well as including also the
imbalance size as an additional weight.
An important next step is to extend these calculations
to beyond the leading-twist contributions, both in the
kinematic factors and the DiFFs themselves. The need
for this is motivated by the upcoming and planned next
generation experiments.
ACKNOWLEDGEMENTS
The work of H.H.M. and A.W.T. was supported by
the Australian Research Council through the ARC Cen-
tre of Excellence for Particle Physics at the Terascale
(CE110001104), and by the ARC Discovery Project No.
DP151103101, as well as by the University of Adelaide.
A.K. was supported by A.I. Alikhanyan National Sci-
ence Laboratory (YerPhI) Foundation, Yerevan, Arme-
nia. The work of A.B. and M.R. is supported by the
European Research Council (ERC) under the European
Union’s Horizon 2020 research and innovation program
(grant agreement No. 647981, 3DSPIN). A. Courtoy
wishes to thank the CICOPS of the University of Pavia
for the support received during the preparation of this
article.
 12

Appendix A

By performing all the integrations indicated in Eq. (46) except for the one upon dqT , the hcos(ϕR − ϕR̄ )i moment
becomes

3α2 X 2
Z Z Z Z
hcos(ϕR −ϕR̄ )i(qT ) = e ¯
dξ dξ d kT d2 k̄T δ(kT + k̄T − qT )
2
(A1)
πQ2 a,ā a
Z X cos n(ϕk − ϕR ) cos m(ϕ − ϕ )
k̄ R̄
× dϕR dϕR̄ cos(ϕR − ϕR̄ )
n,m
π(1 + δ 0,n ) π(1 + δ 0,m )
(
a,[n] ā,[m]
× A(y)D1 D̄1

|kT | ⊥a,[n] |k̄T | ⊥ā,[m] |RT | ^a,[n] |R̄T | ^ā,[m]

+ B(y) cos(ϕk + ϕk̄ ) H H̄ + B(y) cos(ϕR + ϕR̄ ) H H̄
Mh 1 M̄h 1 Mh 1 M̄h 1
|kT | ⊥a,[n] |R̄T | ^ā,[m] |RT | ^a,[n] |k̄T | ⊥ā,[m]
+ B(y) cos(ϕk + ϕR̄ ) H H̄ + B(y) cos(ϕk̄ + ϕR ) H H̄
Mh 1 M̄h 1 Mh 1 M̄h 1
)
|kT ||RT | ⊥a,[n] |k̄T ||R̄T | ⊥ā,[m]
− A(y) sin(ϕk − ϕR ) sin(ϕk̄ − ϕR̄ ) G1 Ḡ1
Mh2 M̄h2

(
3α2 X 2
Z Z Z Z
a,[1] ā,[1]
= e 2 2
d kT d k̄T δ(kT + k̄T − qT ) A(y) cos(ϕk − ϕk̄ ) dξD1 dξ¯ D̄1
πQ2 a,ā a
|kT | ⊥a,[1] |k̄T | ⊥ā,[1]
Z Z
+ B(y) cos(ϕk − ϕk̄ ) cos(ϕk + ϕk̄ ) dξ H1 dξ¯ H̄
Mh M̄h 1
"
|RT | ^a,[2] |R̄T | ^ā,[0] |RT | ^a,[0] |R̄T | ^ā,[2]
Z Z Z Z
1
+ B(y) cos 2ϕk dξ H1 dξ¯ H̄1 + cos 2ϕk̄ dξ H1 dξ¯ H̄
2 Mh M̄h Mh M̄h 1
|kT | ⊥a,[1] |R̄T | 
Z Z 
^ā,[0] ^ā,[2]
+ dξ H1 dξ¯ cos 2ϕk H̄1 + cos 2ϕk̄ H̄1
Mh M̄h
#
| | |R |
Z Z
k̄ T ⊥ā,[1] T

^a,[0] ^a,[2]

+ dξ¯ H̄ dξ cos 2ϕk̄ H1 + cos 2ϕk H1
M̄h 1 Mh
)
|kT ||RT |  ⊥a,[0] |k̄T ||R̄T |  ⊥ā,[0]
Z Z
1 ⊥a,[2] ¯ ⊥ā,[2]

− A(y) cos(ϕk − ϕk̄ ) dξ G1 − G1 dξ Ḡ1 − Ḡ1
4 Mh2 M̄h2
3α2 X 2
Z Z
2
≡ e d kT d2 k̄T δ(kT + k̄T − qT )
πQ2 a,ā a
( )
a a B(y) a B(y) a
× A(y) cos(ϕk − ϕk̄ )F1 + B(y) cos(ϕk − ϕk̄ ) cos(ϕk + ϕk̄ )F2 + cos 2ϕk F3 + cos 2ϕk̄ F4 ,
2 2

where
 13

Z Z
a,[1] ā,[1]
F1a (z, z̄, kT2 , k̄T2 , RT2 , R̄T2 ) = dξD1 dξ¯D̄1 (A2)

|kT ||RT |  ⊥a,[0] ¯|k̄T ||R̄T | Ḡ⊥ā,[0] − Ḡ⊥ā,[2]

Z Z
1 ⊥a,[2]
 
− dξ G 1 − G 1 dξ 1 1
4 Mh2 M̄h2
h i
≡ |kT ||k̄T | F1D (z, kT2 , RT2 )F̄1D̄ (z̄, k̄T2 , R̄T2 ) + F1G (z, kT2 , RT2 )F̄1Ḡ (z̄, k̄T2 , R̄T2 )
|kT | ⊥a,[1] |k̄T | ⊥ā,[1]
Z Z
a 2 2 2 2
F2 (z, z̄, kT , k̄T , RT , R̄T ) = dξ H1 dξ¯ H̄ ≡ |kT ||k̄T |F2H (z, kT2 , RT2 )F̄2H̄ (z̄, k̄T2 , R̄T2 ) (A3)
Mh M̄h 1
|RT | ^a,[2]  |R̄ | |k̄T | ⊥ā,[1]  |kT | ⊥a,[1] |R̄T | ^ā,[0]
Z Z Z Z
T ^ā,[0]
F3a (z, z̄, kT2 , k̄T2 , RT2 , R̄T2 ) = dξ H1 dξ¯ H̄1 + H̄1 + dξ H1 dξ¯ H̄
Mh M̄h M̄h M h M̄h 1
(A4)
|R̄T | ^ā,[2] |RT | ^a,[0] |kT | ⊥a,[1] |k̄T | ⊥ā,[1] |RT | ^a,[0]
Z Z   Z Z
F4a (z, z̄, kT2 , k̄T2 , RT2 , R̄T2 ) = dξ¯ H̄1 dξ H1 + H1 + dξ¯ H̄1 dξ H .
M̄h M h M h M̄h Mh 1
(A5)

The integral on dϕq of the hcos(ϕR − ϕR̄ )i(qT ) moment in Eq. (A1) is nonzero. In fact, the first term of the last
line gives

Z Z Z
dϕq d2 kT d2 k̄T δ(kT + k̄T − qT ) cos(ϕk − ϕk̄ ) F1a (z, z̄, kT2 , k̄T2 , RT2 , R̄T2 ) (A6)
Z Z Z Z
1 ib·(qT −kT −k̄T )
 
= dϕq d2 kT d2 k̄T d2
b e k T · k̄ T F1
D D̄
F̄1 + F G Ḡ
1 F̄ 1
(2π)2
Z Z " #
1 ∂ ∂ ∂ ∂
= dϕq d2 b eib·qT (−4b2 ) F̂ D (z, b2 , RT2 ) 2 F̄ˆ1D̄ (z̄, b2 , R̄T2 ) + 2 F̂1G (z, b2 , RT2 ) 2 F̄ˆ1Ḡ (z̄, b2 , R̄T2 )
(2π)2 ∂b2 1 ∂b ∂b ∂b
Z
= dϕq F1a (qT2 , z, z̄, RT2 , R̄T2 ) = 2πF1a 6= 0,

where F̂1D , F̂1G (F̄ˆ1D̄ , F̄ˆ1Ḡ ) are the inverse Fourier transforms of F1D , F1G (F̄1D̄ , F̄1Ḡ ), respectively.
Following similar steps, it is easy to verify that

Z Z Z
dϕq d2 kT d2 k̄T δ(kT + k̄T − qT ) cos(ϕk − ϕk̄ ) cos(ϕk + ϕk̄ ) F2a (z, z̄, kT2 , k̄T2 , RT2 , R̄T2 ) (A7)
Z Z Z Z
1
2
= dϕq d kT d k̄T 2
d2 b eib·(qT −kT −k̄T ) kT · k̄T cos(ϕk + ϕk̄ ) F2H F̄2H̄
(2π)2
∂2
Z n
= 32 dϕq (qy2 − qx2 ) F a (q 2 , z, z̄, RT2 , R̄T2 )
∂(qT2 )2 2 T
h ∂3 ∂ 4 i 0a 2 o
+ 8 3(qx2 − qy2 ) + (q 4
x − q 4
y ) F (q , z, z̄, R 2
, R̄ 2
) = 0,
∂(qT2 )3 ∂(qT2 )4 2 T T T

Z Z Z
dϕq d2 kT d2 k̄T δ(kT + k̄T − qT ) cos 2ϕk F3a (z, z̄, kT2 , k̄T2 , RT2 , R̄T2 ) (A8)

∂2
Z
= 16 dϕq (qx2 − qy2 ) F a (q 2 , z, z̄, RT2 , R̄T2 ) = 0,
∂(qT2 )2 3 T

and similarly for F4a (z, z̄, kT2 , k̄T2 , RT2 , R̄T2 ).

[1] Vincenzo Barone, Alessandro Drago, and Philip G. Rat-
cliffe, “Transverse polarisation of quarks in hadrons,”
Phys. Rept. 359, 1–168 (2002), arXiv:hep-ph/0104283.
[2] Zhong-Bo Kang, Alexei Prokudin, Peng Sun, and Feng
Yuan, “Extraction of Quark Transversity Distribution
and Collins Fragmentation Functions with QCD Evolu-
tion,” Phys. Rev. D93, 014009 (2016), arXiv:1505.05589
[hep-ph].
[3] M. Anselmino, M. Boglione, U. D’Alesio, J. O. Gonza-
lez Hernandez, S. Melis, F. Murgia, and A. Prokudin,
“Collins functions for pions from SIDIS and new
e+ e− data: a first glance at their transverse mo-
mentum dependence,” Phys. Rev. D92, 114023 (2015),
arXiv:1510.05389 [hep-ph].
[4] Alessandro Bacchetta, Aurore Courtoy, and Marco
Radici, “First glances at the transversity parton dis-
tribution through dihadron fragmentation functions,”
Phys.Rev.Lett. 107, 012001 (2011), arXiv:1104.3855
[hep-ph].
[5] Silvia Pisano and Marco Radici, “Di-hadron fragmenta-
tion and mapping of the nucleon structure,” Eur. Phys.
J. A52, 155 (2016), arXiv:1511.03220 [hep-ph].
[6] John C. Collins, “Fragmentation of transversely polar-
ized quarks probed in transverse momentum distribu-
tions,” Nucl. Phys. B396, 161–182 (1993), arXiv:hep-
ph/9208213 [hep-ph].
[7] P.J. Mulders and R.D. Tangerman, “The Complete tree
level result up to order 1/Q for polarized deep inelas-
tic leptoproduction,” Nucl.Phys. B461, 197–237 (1996),
arXiv:hep-ph/9510301 [hep-ph].
[8] Daniel Boer, R. Jakob, and P.J. Mulders, “Asymme-
tries in polarized hadron production in e+ e- annihila-
tion up to order 1/Q,” Nucl.Phys. B504, 345–380 (1997),
arXiv:hep-ph/9702281 [hep-ph].
[9] A. Bianconi, S. Boffi, R. Jakob, and M. Radici, “Two
hadron interference fragmentation functions. Part 1.
General framework,” Phys.Rev. D62, 034008 (2000),
arXiv:hep-ph/9907475 [hep-ph].
[10] A. Bianconi, S. Boffi, R. Jakob, and M. Radici, “Two
hadron interference fragmentation functions. Part 2.
A Model calculation,” Phys.Rev. D62, 034009 (2000),
arXiv:hep-ph/9907488 [hep-ph].
[11] Marco Radici, Rainer Jakob, and Andrea Bianconi,
“Accessing transversity with interference fragmentation
functions,” Phys.Rev. D65, 074031 (2002), arXiv:hep-
ph/0110252 [hep-ph].
[12] Alessandro Bacchetta and Marco Radici, “Partial wave
analysis of two hadron fragmentation functions,” Phys.
Rev. D67, 094002 (2003), arXiv:hep-ph/0212300 [hep-
ph].
[13] Daniel Boer, Rainer Jakob, and Marco Radici, “Inter-
ference fragmentation functions in electron positron an-
nihilation,” Phys.Rev. D67, 094003 (2003), arXiv:hep-
ph/0302232 [hep-ph].
[14] Alessandro Bacchetta, Federico Alberto Ceccopieri, As-
mita Mukherjee, and Marco Radici, “Asymmetries in-
volving dihadron fragmentation functions: from DIS
to e+e- annihilation,” Phys.Rev. D79, 034029 (2009),
arXiv:0812.0611 [hep-ph].
[15] Aurore Courtoy, Alessandro Bacchetta, Marco Radici,
and Andrea Bianconi, “First extraction of Interference
14
Fragmentation Functions from e+ e− data,” Phys.Rev.
D85, 114023 (2012), arXiv:1202.0323 [hep-ph].
[16] Alessandro Bacchetta and Marco Radici, “Dihadron
interference fragmentation functions in proton-proton
collisions,” Phys.Rev. D70, 094032 (2004), arXiv:hep-
ph/0409174 [hep-ph].
[17] Marco Radici, Alessandro M. Ricci, Alessandro Bac-
chetta, and Asmita Mukherjee, “Exploring universality
of transversity in proton-proton collisions,” Phys. Rev.
D94, 034012 (2016), arXiv:1604.06585 [hep-ph].
[18] Federico A. Ceccopieri, Marco Radici, and Alessan-
dro Bacchetta, “Evolution equations for extended di-
hadron fragmentation functions,” Phys.Lett. B650, 81–
89 (2007), arXiv:hep-ph/0703265 [HEP-PH].
[19] John Collins, Foundations of perturbative QCD (Cam-
bridge University Press, 2013).
[20] A. Airapetian et al. (HERMES Collaboration), “Evi-
dence for a Transverse Single-Spin Asymmetry in Lep-
toproduction of pi+pi- Pairs,” JHEP 06, 017 (2008),
arXiv:0803.2367 [hep-ex].
[21] C. Adolph et al. (COMPASS Collaboration), “Trans-
verse spin effects in hadron-pair production from semi-
inclusive deep inelastic scattering,” Phys.Lett. B713, 10–
16 (2012), arXiv:1202.6150 [hep-ex].
[22] C. Adolph et al. (COMPASS Collaboration), “A high-
statistics measurement of transverse spin effects in
dihadron production from muon-proton semi-inclusive
deep-inelastic scattering,” Phys.Lett. B736, 124–131
(2014), arXiv:1401.7873 [hep-ex].
[23] A. Vossen et al. (Belle Collaboration), “Observation
of transverse polarization asymmetries√ of charged pion
pairs in e+ e− annihilation near s = 10.58 GeV,”
Phys.Rev.Lett. 107, 072004 (2011), arXiv:1104.2425
[hep-ex].
[24] R. Seidl et al. (Belle), “Invariant-mass and fractional-
energy dependence of inclusive
√ production of di-hadrons
in e+ e− annihilation at s = 10.58 GeV,” Phys. Rev.
D96, 032005 (2017), arXiv:1706.08348 [hep-ex].
[25] Marco Radici, A. Courtoy, Alessandro Bacchetta,
and Marco Guagnelli, “Improved extraction of valence
transversity distributions from inclusive dihadron pro-
duction,” JHEP 05, 123 (2015), arXiv:1503.03495 [hep-
ph].
[26] Alessandro Bacchetta, A. Courtoy, and Marco Radici,
“First extraction of valence transversities in a collinear
framework,” JHEP 1303, 119 (2013), arXiv:1212.3568
[hep-ph].
[27] L. Adamczyk et al. (STAR), “Observation of Transverse
Spin-Dependent Azimuthal
√ Correlations of Charged Pion
Pairs in p↑ + p at s = 200 GeV,” Phys. Rev. Lett. 115,
242501 (2015), arXiv:1504.00415 [hep-ex].
[28] Joseph Adams et al. (STAR), “Transverse spin-
dependent azimuthal correlations √ of charged pion pairs
measured in p↑ +p collisions at s = 500 GeV,” (2017),
arXiv:1710.10215 [hep-ex].
[29] Marco Radici, “Update on extraction of transversity PDF
from inclusive di-hadron production,” in 25th Interna-
tional Workshop on Deep Inelastic Scattering and Related
Topics (DIS 2017) Birmingham, UK, April 3-7, 2017 ,
Vol. 229 (2017) arXiv:1709.00360 [hep-ph].
[30] W. Bentz, A. Kotzinian, H. H. Matevosyan, Y. Ninomiya,

A. W. Thomas, and K. Yazaki, “Quark-Jet model for
transverse momentum dependent fragmentation func-
tions,” Phys. Rev. D94, 034004 (2016), arXiv:1603.08333
[nucl-th].
[31] Hrayr H. Matevosyan, Aram Kotzinian, and An-
thony W. Thomas, “Monte Carlo Implementation of Po-
larized Hadronization,” Phys. Rev. D95, 014021 (2017),
arXiv:1610.05624 [hep-ph].
[32] Hrayr H. Matevosyan, Aram Kotzinian, and An-
thony W. Thomas, “Dihadron fragmentation functions in
the quark-jet model: Longitudinally polarized quarks,”
Phys. Rev. D96, 074010 (2017), arXiv:1707.04999 [hep-
ph].
[33] Hrayr H. Matevosyan, Aram Kotzinian, and An-
thony W. Thomas, “Dihadron fragmentation functions
in the quark-jet model: Transversely polarized quarks,”
Phys. Rev. D97, 014019 (2018), arXiv:1709.08643 [hep-
ph].
[34] Alessandro Bacchetta and Marco Radici, “Two hadron
semiinclusive production including subleading twist,”
Phys.Rev. D69, 074026 (2004), arXiv:hep-ph/0311173
[hep-ph].
[35] A. Abdesselam et al. (Belle), “Measurement of Azimuthal
Modulations in the Cross-Section of Di-Pion Pairs in
Di-Jet Production from Electron-Positron Annihilation,”
(2015), arXiv:1505.08020 [hep-ex].
[36] Anselm Vossen, “Recent Fragmentation Function Mea-
surements at Belle,” Proceedings, 23rd International
15
Workshop on Deep-Inelastic Scattering and Related Sub-
jects (DIS 2015): Dallas, Texas, USA, April 27-May 01,
2015, PoS DIS2015, 216 (2015).
[37] Stefan Sirtl, “Azimuthal asymmetries in SIDIS di-hadron
muoproduction off longitudinally polarized protons at
COMPASS,” in 22nd International Symposium on Spin
Physics (SPIN 2016) Urbana, IL, USA, September 25-30,
2016 (2017) arXiv:1702.07317 [hep-ex].
[38] Daniel Boer, “Angular dependences in inclusive two-
hadron production at BELLE,” Nucl. Phys. B806, 23–67
(2009), arXiv:0804.2408 [hep-ph].
[39] Stephen Gliske, Alessandro Bacchetta, and Marco
Radici, “Production of two hadrons in semi-inclusive
deep inelastic scattering,” Phys. Rev. D90, 114027
(2014), [Erratum: Phys. Rev.D91,no.1,019902(2015)],
arXiv:1408.5721 [hep-ph].
[40] Daniel Boer, R. Jakob, and P. J. Mulders, “Leading
asymmetries in two hadron production in e+ e- annihi-
lation at the Z pole,” Phys. Lett. B424, 143–151 (1998),
arXiv:hep-ph/9711488 [hep-ph].
[41] Hrayr H. Matevosyan, Aram Kotzinian, and Anthony W.
Thomas, “Accessing quark helicity through dihadron
studies,” (2017), arXiv:1712.06384 [hep-ph].
[42] Anselm Vossen, “Modulations of the di-pion pair cross
section at belle,” (2013), talk at the Indiana–Illinois
workshop on fragmentation functions, Bloomington (IN-
USA) 12 - 14 Dec 2013.