PHYSICAL REVIEW C 81, 067601 (2010)

Phase ambiguity of the threshold amplitude in pp pp

0
G. Ramachandran,
1
G. Padmanabha,
1,2
and Sujith Thomas
1,3
1
G. V. K. Academy, Jayanagar, Bangalore 560082, India
2
Department of Physics, Bangalore University, Bangalore 560056, India
3
K. S. Institute of Technology, Bangalore 560056, India
(Received 25 March 2010; published 14 June 2010)
Measurements of spin observables in pp p p
0
are suggested to remove the phase ambiguity of the
threshold amplitude. The suggested measurements complement the Indiana University Cyclotron Facility data
on p p pp
0
to completely determine all the 12 partial wave amplitudes taken into consideration by Meyer
et al. [Phys. Rev. C 63, 064002 (2001)] and Deepak et al. [Phys. Rev. C 72, 024004 (2005)].
DOI: 10.1103/PhysRevC.81.067601 PACS number(s): 25.40.Ve, 24.70.+s, 13.75.Cs, 13.75.Gx
Introduction. Meson production in NN collisions has con-
tinued to excite considerable interest [14] since total cross-
section measurements [5] for pp pp
0
in the early 1990s
were found to be more than a factor of 5 larger than the then
available theoretical predictions [6]. To bridge the gap between
experiment and theory, several mechanisms, like exchange of
heavy mesons, two-pion exchange, off-shell extrapolation of
the vertex form factor, nal-state interactions, and contribu-
tions owing to resonance and of low-lying nucleon reso-
nances, were proposed. Hanhart et al. [7] in 2000 observed:
As far as microscopic model calculations of the reaction
NN NN are concerned, one has to concede that theory is
denitely lagging behind the development of the experimental
sector . . . Further more they take into account only the lowest
partial wave(s). Therefore, it is not possible to confront these
models with the wealth of experimental information available
nowadays specically with differential cross-sections and with
spin dependent observables. The Julich model, on the other
hand, takes into consideration higher partial waves as well.
In contrast to elastic NN scattering, where channel
spin is conserved, the pp pp
0
transition at threshold
to the nal Ss state is a triplet to singlet. Next in order
are the transitions to Ps states, which are singlet to triplet.
As the energy is increased, transitions to Pp states are also
expected to contribute, which are, however, triplet to triplet.
Pionic d-wave effects were reported [8] even at a beam energy
of 310 MeV. Measurements up to 425 MeV have also been
reported [9], where evidence for a Ds state was seen even
at 310 MeV. Advances in storage ring technology [10] led
to detailed experimental studies, including measurements of
spin observables employing polarized beams of protons on
polarized proton targets. Of the two existing models [11,12]
which include higher partial waves, the Julich meson exchange
model [7,11] was thoroughly confronted with these data. The
model was comparatively more successful with the less com-
plete data on p p d
+
[13] and p p pn
+
[14] but failed
to provide an overall satisfactory reproduction of the complete
set of polarization observables in the case of p p pp
0
[15]. In this context, a model-independent approach [16,17]
was developed using irreducible tensor techniques [18]. The
reaction is characterized, in this formalism, by irreductible
tensor amplitudes M

(s
f
, s
i
) of rank = |s
f
s
i
|, . . . , (s
f
+
s
i
), where s
i
, s
f
denote the initial and nal spin states of the
two protons. Each of these amplitudes is expressible in terms
of partial wave amplitudes M
j
l(l
f
s
f
)j
f
;l
i
s
i
, which are functions
of the c.m. energy E and invariant mass W of the two-proton
system in the nal state. The relative orbital angular momenta
between the two protons in the initial and nal states are
denoted by l
i
and l
f
, respectively, and l denotes the pion orbital
angular momentumin the c.m. frame. The threshold amplitude
M
0
0(00)0;11
contributes to M
1
0
(0, 1), and an empirical estimate
of the integrated |M
1
0
(0, 1)|
2
was presented in Ref. [16],
based on the then existing data [5]. The same approach was
employed subsequently to analyze [19] the Indiana University
Cyclotron Facility data on p p pp
0
[15] immediately after
its publication. The 16 partial waves listed by Meyer et al. [15]
covered the Ss, Ps, Pp, Sd, and Ds channels. Here, the capital
letters denote l
f
while the lower case indicate l. In Ref. [20],
the same set of partial waves were listed, of which, the last
four, covering Sd and Ds, were ignored following Ref. [15].
Because the nal spin-singlet and spin-triplet states do not
mix in any of the spin observables measured in Ref. [15], the
Ss amplitude and the larger of the Ps amplitudes were both
chosen to be real in Ref. [20]. This implies that the phase of
the Ss amplitude remained ambiguous but chosen to be zero
with respect to the larger Ps amplitude. The comparison of
the empirically extracted amplitudes with the Julich model
predictions revealed that (i) the contributions are important
and (ii) the model deviated very strongly in the case of
3
P
1

3
P
0
p and to a lesser extent in
3
F
3

3
P
2
p, which
will guide the search for the possible shortcomings [20].
The purpose of the present Brief Report is to extend
the model-independent theoretical discussion to the spin
polarization of the protons in the nal state and to examine how
the additional experimental measurements regarding the nal
spin state can be used to determine empirically the strengths
of all these amplitudes and the ambiguous relative phase of the
threshold Ss amplitude with respect to the 11 near-threshold
Ps and Pp amplitudes considered in Refs. [15,19,20] for
pp pp
0
. We mention that the p-wave charged pion
production in pn pp

, pp pn
+
, pp d
+
has
more recently been discussed [21] using effective eld
theory and it was proposed earlier [4,12] to pin down the
phase of the Ss partial wave amplitude with reference to
the isospin-0-to-isospin-1 Sp amplitude by looking at the
forward-backward assymetry in pn pp

.
0556-2813/2010/81(6)/067601(4) 067601-1 2010 The American Physical Society
BRIEF REPORTS PHYSICAL REVIEW C 81, 067601 (2010)
Theoretical formalism. We consider the reaction pp
pp
0
at c.m. energy E and initial c.m. momentum p
i
= p
i
p
i
,
which may be chosen to be along the z axis. Let q = q q =
(p
1
+p
2
) denote the pion momentum in the c.m. frame and
let p
f
= p
f
p
f
=
1
2
(p
1
p
2
) in terms of the c.m. momenta p
1
and p
2
of the two protons in the nal state.
Following Ref. [16], we write the matrix M in spin space
for the reaction pp pp
0
in the form
M =
1

s
i
,s
f
=0
s
i
+s
f

=|s
i
s
f
|
(S

(s
f
, s
i
) M

(s
f
, s
i
)), (1)
where s
i
and s
f
denote the initial and nal channel spins,
respectively. The irreducible tensor operators S

(s
f
, s
i
) of rank
, with taking values = , 1, . . . , , are dened
in Ref. [18]. The irreducible tensor amplitudes M

(s
f
, s
i
) in
Eq. (1) are expressible as
M

(s
f
, s
i
) =

L,j
W(l
i
s
i
L
f
s
f
; j) Z(s
f
, s
i
, L, j)A

(L),
(2)
where
A

(L) =
__
Y
l
f
( p
f
) Y
l
( q)
_
L
f
Y
l
i
( p
i
)
_

, (3)
and the symbol L is used to collectively denote L
{l
f
, l, L
f
, l
i
}. It may be noted that (1)
l
f
+l+l
i
= 1 owing
to parity conservation. The complex numbers Z(s
f
, s
i
, L, j)
are given by
Z(s
f
, s
i
, L, j) =
[L
f
][j]
2
[s
f
]
(1)
js
i
+1

j
f
[j
f
]W(s
f
l
f
jl; j
f
L
f
)M
j
l(l
f
s
f
)j
f
;l
i
s
i
(4)
in terms of the 16 partial wave reaction amplitudes
M
j
l(l
f
s
f
)j
f
;l
i
s
i
= F (l(l
f
s
f
)j
f
)j||T ||(l
i
s
i
)j, (5)
proportional to the reduced on-energy-shell T -matrix elements
(l(l
f
s
f
)j
f
)j||T ||(l
i
s
i
)j for the reaction. The purely kinemat-
ical factor
F = (i)
l
i
ll
f
4(2)
1/2
_
W(E )qp
f
/p
i
(6)
is introduced explicitly in Eq. (5) so that the dependence
on E and W is seen to be completely taken care of by the
M
j
l(l
f
s
f
)j
f
;l
i
s
i
. They are identical to the amplitudes denoted as
T in Ref. [20]. We may, following Refs. [15,20], neglect the
last 4 amplitudes, which are Sd and Ds, and consider the
rst 12 amplitudes, which are, for simplicity, enumerated as
f
1
, . . . , f
12
in Table I.
The unpolarized double-differential cross section may now
be written as
d
2

0
dW d
f
d
=
1
4
Tr[M M

], (7)
where M

denotes the Hermitian conjugate of M given by

Eq. (1). The invariant mass W of the two protons in the nal
TABLE I. List of the partial wave amplitudes for the reaction
pp pp
0
.
Initial pp Type Final pp
0
Partial wave
state state amplitudes
3
P
0
Ss
1
S
0
, s M
0
0(00)0;11
= f
1
1
S
0
Ps
3
P
0
, s M
0
0(11)0;00
= f
2
1
D
2
3
P
2
, s M
2
0(11)2;20
= f
3
3
P
0
Pp
3
P
1
, p M
0
1(11)1;11
= f
4
3
P
2
3
P
1
, p M
2
1(11)1;11
= f
5
3
P
2
3
P
2
, p M
2
1(11)2;11
= f
6
3
F
2
3
P
1
, p M
2
1(11)1;31
= f
7
3
F
2
3
P
2
, p M
2
1(11)2;31
= f
8
3
P
1
3
P
0
, p M
1
1(11)0;11
= f
9
3
P
1
3
P
1
, p M
1
1(11)1;11
= f
10
3
P
1
3
P
2
, p M
1
1(11)2;11
= f
11
3
F
3
3
P
2
, p M
3
1(11)2;31
= f
12
state is given by
W =
_
_
E
2
+m
2

2E
_
, (8)
where m

denotes the pion mass and denotes the c.m. energy

of pion. It may be noted that
d
2

0
d
3
p
f
d
=
W
p
f
d
2

0
dWd
f
d
. (9)
It is worth noting that the threshold Ss amplitude f
1
alone
contributes to
M
1

(0, 1) =
1
4

3
f
1
Y
1
( p
i
), (10)
which is spherically symmetric both with respect to p
f
as well
as to q in the nal state, while all the other irreducible tensor
amplitudes are independent of f
1
.
Final-state polarization with initially unpolarized protons.
If the colliding protons are unpolarized, the spin density matrix

f
characterizing the two protons in the nal state is given by

f
=
1
4
MM

, (11)
so that Eq. (7) is identical to Tr[
f
].
The nal spin state is completely determined through
measurements of the polarizations
P
i
=
Tr[
i

f
]
Tr[
f
]
, i = 1, 2 (12)
of the two protons and their spin correlations
C

=
Tr[
1

2

f
]
Tr[
f
]
, , = x, y, z. (13)
All these spin observables may elegantly be calculated by
considering
P
k

(s

f
, s
f
) = Tr
_
S
k

(s

f
, s
f
)
f
_
, (14)
067601-2
BRIEF REPORTS PHYSICAL REVIEW C 81, 067601 (2010)
where S
k

(s

f
, s
f
) are given in terms of the Pauli spin matrices

1
and
2
of the two protons in the nal state through
S
0
0
(0, 0) =
1
4
(1
1

2
), (15)
S
0
0
(1, 1) =
1
4
(3 +
1

2
), (16)
S
1

(1, 1) =

3
2

2
(
1
+
2
)
1

, (17)
S
2

(1, 1) =

3
2
(
1

2
)
2

, (18)
S
1

(0, 1) =
1
2

2
(
1

2
)
1

1
4
(
1

2
)
1

, (19)
S
1

(1, 0) =

3
2

2
(
1

2
)
1

3
4
(
1

2
)
1

. (20)
Thus, the double-differential cross section is given by
d
2

0
dW d
f
d
= Tr[
f
] = P
0
0
(0, 0) +P
0
0
(1, 1), (21)
in terms of the double-differential cross sections, P
0
0
(0, 0)
leading to the nal singlet state and P
0
0
(1, 1) leading to the
nal triplet state of the two protons. If we use the notations
(P
i
)

to denote the spherical components, that is,

(P
i
)
0
= P
iz
; (P
i
)
1
=
1

2
(P
ix
P
iy
), (22)
it follows from Eqs. (19) and (20) that
P
1

(1, 0)

3P
1

(0, 1) =

3
2
Tr[
f
](P
1
P
2
)

, (23)
whereas it follows from Eq. (17) that
P
1

(1, 1) =

3
2

2
Tr[
f
](P
1
+P
2
)

, (24)
which together determine P
1
and P
2
individually. Finally, the
spin correlations C

dened in Eq. (13) may likewise be

related to Eq. (14) using
P
0
0
(1, 1) 3P
0
0
(0, 0) = Tr[(
1

2
)
f
], (25)
P
1

(1, 0) +

3P
1

(0, 1) =

3i
2
Tr[
f
(
1

2
)]

, (26)
P
2

(1, 1) =

3
2
Tr
_

f
(
1

2
)
2

_
. (27)
Using the known properties [18] of the spin operators S

and
standard Racah techniques, we may obtain a master formula
for all the nal-state spin observables, which is given by
P
k

(s

f
, s
f
) =
1
4

s
i
,,

(1)
s
f
s
i
[s
f
] [s

f
]
2
[][

]
W(s

s
f
; s
i
k)[M

(s
f
, s
i
) M

(s

f
, s
i
)]
k

,
(28)
where M

(s
f
, s
i
) are dened in terms of the complex
conjugates M

(s
f
, s
i
)

of M

(s
f
, s
i
) given by Eq. (2) through
M

(s
f
, s
i
) = (1)

(s
f
, s
i
)

. (29)
Noting once again that (1)
l
f
+l+l
i
= 1, owing to parity
conservation, we may express
M

(s
f
, s
i
) = (1)
1

L
W(l
i
s
i
L
f
s
f
; j)
Z

(s
f
, s
i
, j, L) A

(L), (30)
where Z

(s
f
, s
i
, j, L) denote the complex conjugates of
Z(s
f
, s
i
, j, L) given by Eq. (4).
Relative phase of the threshold amplitude. We may now
take advantage of the fact that M
1
0
(0, 1) given by Eq. (10) is
spherically symmetric with respect to p
f
and q and involves
only the threshold amplitude f
1
. Moreover, M

(1, 1) are
independent of f
1
and depend only on the Pp amplitudes
f
4
, . . . , f
12
. Therefore, we focus attention on Eqs. (26) and
(23), which involve
[M

(1, 1) M

1
(0, 1)]
1

L,j
Z(1, 1, j, L)f

1
A
1

(31)
[M
1
(0, 1) M

(1, 1)]
1

L,j
Z

(1, 1, j, L)f
1
A
1

, (32)
where
A
1

=
1
4

3
W(l
i
1L
f
1; j)[A

(L) Y
1
( p
i
)]
1

. (33)
Expressing
[A

(L) Y
1
( p
i
)]
1

3
4

L
i
W(L
f
l
i
11; L
i
)[][L
i
][l
i
]
C(l
i
1L
i
, 000)A
1

(l
f
lL
f
L
i
) (34)
and carrying out the summation over L and j, we obtain
P
1

(1, 0) = f

1
[F
1
A
1

(1110) +F
2
A
1

(1112) +F
3
A
1

(1122)],
(35)
P
1

(0, 1) =
1

3
f
1
_
F

1
A
1

(1110) +F

2
A
1

(1112)
+F

3
A
1

(1122)
_
, (36)
where F
i
, i = 1, 2, 3 are well-dened linear combinations of
the Pp amplitudes given by
F
1
=
1
32
3/2
_
f
4

5
6
f
5
+
5
2

3
f
6
+
1
3

3
f
9

1
6
f
10

5
6

3
f
11
_
, (37)
F
2
=
1
32

2
3/2
_
f
5

3f
6
+
_
3
2
f
7
+
3

2
f
8
_
, (38)
F
3
=
1
32

2
3/2
_

3f
5
+f
6
+

7f
7
+
_
7
3
f
8
_
, (39)
067601-3
BRIEF REPORTS PHYSICAL REVIEW C 81, 067601 (2010)
Because the Pp amplitudes have been determined both in
magnitude and in relative phase with respect to f
2
in Ref. [20],
we may express F

= |F

| exp[i

] , = 1, 2, 3 and treat
|F

| and

as known. In Ref. [20], f

2
was assumed to be
real. Because the relative phase between f
1
and f
2
could not
be ascertained from the measurements of Meyer et al. [15],
f
1
was also assumed to be real, although only one of the
amplitudes can be taken as real. Therefore, we choose f
2
to be
real and express f
1
= |f
1
| exp[i
1
]. This leads to
P
1

(1, 0)

3P
1

(0, 1) = 2
3

=1
R

cos(

1
)A
1

(),
(40)
P
1

(1, 0) +

3P
1

(0, 1) = 2i
3

=1
R

sin(

1
)A
1

(),
(41)
where R

= |F

||f
1
| and A
1

() for = 1, 2, 3 denote
A
1

(1110), A
1

(1112), A
1

(1122), respectively.
It is seen from Eq. (11) that measuring the double-
differential cross section (7) yields Tr[
f
]. Measurements of
(P
1
P
2
)

given by Eq. (23) then lead to empirical determi-

nation of Eq. (40), while measurements of spin correlations
C
xy
C
yx
, C
yz
C
zy
, C
zx
C
xz
, where C

are given by
Eq. (13) lead to empirical determination of Eq. (41) using
Eq. (26).
Thus, we nd that it is possible to determine empirically the
relative phase
1
of f
1
, without any trigonometric ambiguities,
because R

and

are known from Ref. [20]. We therefore

advocate measurement of these pp spin observables in the nal
state, employing simply an unpolarized beam and unpolarized
target initially, to complement the spin observables measured
by Meyer et al. [15], so that the amplitudes f
1
, f
2
, . . . , f
12
may be determined empirically without any phase
ambiguity.
We thank the referee for bringing Refs. [21,22] to our
attention and suggesting that our formalism be applied to
discuss polarization in strangeness production [22], which will
be taken up later.
[1] H. Machner and J. Haidenbauer, J. Phys. G 25, R231 (1999).
[2] P. Moskal, M. Wolke, A. Khoukaz, and W. Oelert, Prog. Part.
Nucl. Phys. 49, 1 (2002).
[3] G. Faldt, T. Johnson, and C. Wilkin, Phys. Scr. T 99, 146 (2002).
[4] C. Hanhart, Phys. Rep. 397, 155 (2004).
[5] H. O. Meyer et al., Phys. Rev. Lett. 65, 2846 (1990); Nucl.
Phys. A 539, 633 (1992); A. Bondar et al., Phys. Lett. B 356, 8
(1995).
[6] D. S. Koltun and A. Reitan, Phys. Rev. 141, 1413 (1966); G. A.
Miller and P. U. Sauer, Phys. Rev. C 44, R1725 (1991).
[7] C. Hanhart, J. Haidenbauer, O. Krehl, and J. Speth, Phys. Rev.
C 61, 064008 (2000).
[8] J. A. Niskanen, Phys. Lett. B 289, 227 (1992); Y. Maeda et al.,
N Newsl. 13, 326 (1997).
[9] R. Bilger et al., Nucl. Phys. A 693, 633 (2001).
[10] H. O. Meyer et al., Phys. Rev. Lett. 81, 3096 (1998); 83, 5439
(1999).
[11] C. Hanhart, J. Haidenbauer, O. Krehl, and J. Speth, Phys. Lett.
B 444, 25 (1998).
[12] Y. Maeda, N. Matsuoka, and K. Tamura, Nucl. Phys. A 684, 392
(2001).
[13] B. v. Przewoski et al., Phys. Rev. C 61, 064604 (2000).
[14] W. W. Daehnick et al., Phys. Rev. C 65, 024003 (2002).
[15] H. O. Meyer et al., Phys. Rev. C 63, 064002 (2001).
[16] G. Ramachandran, P. N. Deepak, and M. S. Vidya, Phys. Rev. C
62, 011001(R) (2000).
[17] G. Ramachandran and P. N. Deepak, Phys. Rev. C63, 051001(R)
(2001).
[18] G. Ramachandran and M. S. Vidya, Phys. Rev. C56, R12 (1997).
[19] P. N. Deepak and G. Ramachandran, Phys. Rev. C 65, 027601
(2002); P. N. Deepak, G. Ramachandran, and C. Hanhart,
Matter. Mater. 21, 138 (2004); P. N. Deepak, C. Hanhart,
G. Ramachandran, and M. S. Vidya, Int. J. Mod. Phys. A 20,
599 (2005).
[20] P. N. Deepak, J. Haidenbauer, and C. Hanhart, Phys. Rev. C 72,
024004 (2005).
[21] V. Baru, E. Epelbaum, J. Haidenbauer, C. Hanhart, A. E.
Kudryavtsev, V. Lensky, and U. G. Meissner, Phys. Rev. C 80,
044003 (2009).
[22] A. Kacharava, F. Rathmann, and C. Wilkin (ANKE Collabora-
tion), Proposal for the Spin Physics fromCOSYto FAIR, Julich,
August 2005 [arXiv:nucl-ex/0511028v1].
067601-4