You are on page 1of 18

Nuclear Physics A367 (1981) 77-94

@ North-Holland Publishing Company

ISOVECTOR AND ISOSCALAR PAIRING CORRELATIONS


IN A SOLVABLE MODEL

J. A. EVANS

School oj’kfathematieal and Physical Sciences, University of Sussex,Brighton, UK t


and
Departamento de Fisica, CNEA, 1429 Buenos Aires, Argentina

and

G. G. DUSSEL *, E. E. MAQUEDA * and R. P. J. PERAZZO *

Departamento de Fisica, CNEA, 1429 Buenos Aires, Argentina

Received 24 February 1981

(Revised 10 April 1981)

Abstract: A nuclear model in which the nucleons interact via an arbitrary combination of isovector and
isoscalar pairing forces is solved by the use of group theory. Collective features of the energy spectra
are studied, together with the transfer of nucleon pairs and a-particles. While pair transfer is strongly
affected by the relative strength of the two interaction channels, E-transfer is relatively insensitive,
but shows slight enhancement when the two interactions have equal strength.

1. Introduction

Pairing correlations have been introduced in nuclear physics to account for a


wide variety of phenomena that range from the odd-even mass difference to the
enhanced population of J = O+ states in two-nucleon transfer reactions. Depending
on the relative value of the matrix element ofthis residual interaction and the single-
particle splitting, normal or superfluid phases may occur. In a realistic situation
other channels of the effective residual two-body interaction may compete with
the pairing giving rise to different types of correlations. The purpose of the present
paper is to study in detail one of these types of competition within an exactly solvable
model.
With this aim we propose a situation in which neutrons and protons move in a
set of degenerate Z-shells and interact via two generalized pairing forces acting only

t Permanent address.
* Fellow of the Consejo National de Investigaciones Cientiticas y Tecnicas, Argentina.
71
78 J. A. Evans et al. 1 Pairing correlations

in L = 0 states, having T = l(S = 0) and T = O(S= 1) and whose relative


intensity is left as a free parameter. We are thus in the position of studying all the
ordinary isovector (T = 1) pairing features of the system in the presence of a dif-
ferent channel that introduces a kind of correlations which interfere destructively
with it.
The isoscalar (T = 0)channel favours the occurrence of a condensate of pairs
of non-identical nucleons, each carrying S = 1. The internal wave function of each
pair is that of a deuteron except for the relative orbital motion depending on the
principal (radial) quantum number that is assigned to the Z-shells considered. The
relative S-state is, nevertheless, always dominant. We may consider the “deuteron
pairing”, as a way of introducing an isoscalar proton-neutron component into the
effective nuclear interaction, the importance of which, in the spectra of odd-odd
nuclei, has been well established 1,2). The isovector pairing force can be solved
exactly for particles moving in several degenerate I- orj-shells using the O(5) classifica-
tion scheme 3-5) and its eigenvalues can be expressed analytically. If non-degenerate
single-particle levels are considered a numerical diagonalisation 6, ‘) is required. A
pure deuteron pairing force can of course be solved with the same classification
scheme by simply changing the roles of T and S, but only in the absence of spin-
orbit forces.
When the isoscalar and isovector components are both active the O(5) classification
scheme is no longer sufficient and the larger symmetry group O(8) must be introduced
to account for the commutations of all the operators involved *). In ref. 8, closed
analytical expressions are provided for the eigen-energies when both components
have the same strength, taking advantage of the fact that in this case the hamiltonian
is proportional to the Casimir operator of the subgroup SU(4). For our more general
situation we make use of the O(8) group operators to write the wave functions and
to construct the hamiltonian matrices that must in general be diagonalised numerical-
ly.
With the exact solutions thus obtained we study the competition between the
isoscalar and the isovector pairing correlations by analysing energy eigenvalues
and two-particle spectroscopic amplitudes. Another point of interest is the behaviour
of a-transfer cross sections whose enhancement has been linked to the occurrence
of a pairing phase transition ‘-lr).
The internal wave function of an a-like cluster in a heavy or medium nucleus has
a large superposition with that of two isovector pairs coupled to total T = 0.It is
‘therefore intuitively reasonable that the occurrence of a pair condensate will enhance
the a-transfer matrix element. It is however also true that the internal wave function
of an cc-cluster has a large superposition with that of two isoscalar pairs with S = 1
coupled to total S = 0. We may thus expect that the matrix element for an cc-transfer
will also be enhanced whenever the S = 1 pairing channel is strong enough to
produce a “deuteron” condensate. These two intuitive pictures, that correspond to
the exchange of roles of 5 and T,are separately correct. In fact, the evaluation 9,
J. A. Evans et al. / Pairing correlations 79

of the a-transfer spectroscopic amplitudes with a pure pairing force shows such
an enhancement in the superfluid limit. The behaviour of these amplitudes when
both channels are simultaneously active was studied in the p-shell 9, showing that
it was almost independent of the relative intensity of the components of the force.
That calculation is however not conclusive because the degeneracy of the p-shell
may be too small to favour the occurrence of collective phenomena.
In the next section we discuss the group theoretical framework in which we con-
struct the many-body wave functions with definite spin, isospin and particle number
and go through the details of the evaluation of the hamiltonian matrix elements
and the two- and four-particle spectroscopic amplitudes. In sect. 3 we discuss the
numerical results, in particular for a total shell degeneracy 52 = 6 which corre-
sponds to a degenerate s-d shell, though some features are discussed for arbitrary
CL?.

2. Theory

In this section we describe briefly the relationship between the physical operators
of the model and the abstract infinitesimal operators of O(8). We show further how,
by exploiting the properties of this group, the energy matrices and other physical
quantities in the model may be calculated. More detailed accounts have appeared
elsewhere *’ ’ 5).
The discussion is restricted to states of (O(8)) seniority zero.

2.1. DESCRIPTION OF THE MODEL

The hamiltonian has an isoscalar pairing interaction in addition to the usual


isovector pairing, acting in a set of degenerate I-shells with total degeneracy fi =
c (2Z+ 1). Thus

H = -(l +x)cP;PP-(l -x)~D,‘D,, (1)


B P

where, for ,D = 0, * 1,

In these expressions UT,mr,m,,mf is the usual creation operator for a nucleon and the
square bracket denotes coupling of orbital angular momentum, spin and isospin,
respectively.
80 J. A. Evans et al. 1 Pairing correlations

The sixteen number conserving operators

‘N
2 = c J21+l[a:a,];;;,

@b)

where
a, , ml1inb, mt = (-)l+l+mi+m,+mtal,_m~ , _mb3-mt,
close under commutation and are the infinitesimal operators of the group U(4)
of unitary transformations in the spin-isospin space. The irreducible representations
of this group, labelled by the Young tableaux [,I] are also irreducible for its uni-
modular subgroup SU(4) obtained by deleting the number operator N. The hamil-
tonian H is diagonal in N, S and T but not in [k] unless x = 0.
The twenty-eight operators comprising those in (2a) and their hermitean conjugates
together with those in (2b) also close under commutation, and are the infinitesimal
operators of O(8), the group of orthogonal transformations in an eight-dimensional
space *).
The standard form of the infinitesimal operators of O(n) is 12)

J,, = -i xp&-xq& , lzLqcp_Fn,


4 P

and for n = 8 we relate these to the operators defined in (2) by assigning q = 1, 2, 3


to isospin, q = 4, 5, 6 to spin and q = 7, 8 to the gauge space in which rotations
are generated by N. Once the assignments for the J,+ 1, q (1 5 q 5 7) have been
made, the remainder follow from the commutation rules. A suitable choice is

521 = To, J43 = woo, 565 = So, J,, = fN-L’,

J,, = T,, J,, = S,, J,, = +(D’,-D:+D_,-D,). (4)


A summary of the correspondence of the physical operators (2) with the O(8)
generators (3) resulting from the assignments in (4) is given in table 1.
By successively reducing n in unit steps in (3) we generate the subgroup chain,

O(8) 2 O(7) I O(6) I O(5) I O(4) I O(3) 2 O(2), (5)


and the representation labels of all these groups provide a complete classification
for the states spanning a representation of O(8). The states classified in this manner
are those of the Gel’fand basis. Explicit formulas for the matrix elements in this
basis of the J,+ l,q operators have been given by Wong 12).
J. A. Evans et al. / Pairing correlations 81

TABLE 1

Correspondence of physical operators with O(8) generators

J 21
J 32
J 13
J 65
J

${(J,, + ~~~~~
t(-J,,~J,,-i~,,+i~,,)
&J,, + iJ,,)
P(-JslT J7z-iJ71+iJsz)

In the present work we are concerned only with the states of seniority zero, which
span the irreducible representation (QOOO). The dimension of this representation
is 13)

d = &Q+5)(Q+4)(52+3)2(Q+2)(Q+l). (6)
It may be shown 12) that the labels for the groups in (5) are, respectively,

(QOOO), (AOO), (~200), w>, W>> &)Y Gu (7)


with Sz 2 /2, 2 1, 2 II, 2 II, 2 A5 2 I&) where the Izi are integers of which only
A6 can be negative. These six integers define the Gel’fand basis in (QOOO).

2.2. EVALUATION OF THE ENERGY MATRIX

The Gel’fand basis is unsuitable for the evaluation of energy matrices because it
gives non-diagonal forms for N, S2 and S,. Physical states IN, S, M,, T, M,, A2)
may be obtained as linear combinations of the Gel’fand states, essentially by diag-
onalising these operators. It follows from eqs. (4) that T = A, and M, = &. The
quantum number 1, enters directly because the isospin-spin symmetry label is
[A,, A,] as the O(6) group in (5) is homomorphic to the SU(4) mentioned earlier.
The problem of obtaining the physical states has been discussed in detail by Pang l 5),
who gives explicit formulas for the matrix elements of the operators in (2a) between
physical states of seniority 0 or 1. Once the reduced matrix elements of Pf, D’,
or spectroscopic amplitudes, are known, the matrix elements of H are given by

= -(l-x)x (N,S, 7;il,lJD+IlN-2,S, ?;&)(N,S, ?;-A;.IID+tlW-2,S, 7;&)


s, nz

(8)
2.3. LIMITING CASES

It is well known that by exploitation of the Casimir i~variants of U(8) and some
of its subgroups,,simple formulas can be obtained for the eigenvalues of H in some
limiting cases **14). Pure isovector pairing arises when x = 1. The relevant sym-
metry group in this case is O(5), the orthogonal group in the subspace g = 1, 2,
3,7,8. [NB This is not the O(5) in (5).] The infinitesimal operators of this group are
P,‘, Pp, T, and N. The irreducible.repres~ntatio~s are labelled by (52- +vl, t> where
D%is the isovector seniority and t the reduced isospin. The ~lass~~~ation with respect
to v1 of all the states of the irreducible representation (6~~~) of O(8) fd = 1386) is
given in table 2. All these states have t = 0. The eigen~alues of H are given by

-$(N-v,)(4ft+S-N-v,)+ T(T+ 1). (9)


A similar result is obtained for pure isoscalar pairing when x = - 1. In this case
the symmetry group is the orthogonal group in the subspace q = 4,5,6,7,8 which
has as infinitesimal operators D:, D,, S,, and N. The irreducible representations
are labelled by the isoscalar seniority q, and the eigenvalues of N are given by
-S(N-~,)(~~Z~~-N--~)+S(SS-I). (IO)
The isoscalar seniority classification for the states of (6~~~) is as shown in table 2
with S and T interchanged.
When x = 0 the symmetry group is SU(4) or O(6), providing the isospin-spin
symmetry label f&, &]. These states occur for N = 2/J,,,21, -I-4,2/1, + 8, . . ., 452- 2&
and tables of the associated S’ and T values are available 16). The eigenvalues of W
are given in this case by

~N2+3~2(122+4)-3N(I(Z+3). (11)
These limiting cases provide very stringent checks on the correctness of the energy
matrices.

2.4. SPECTROSCOPIC A~RL~T~DES FOR a-TRANSFER

The operator for a-particle transfer in the model is

which satisfies the commutation relations

CJw, T,*l = 27X+,


l&p Cl = 0, l~;q<p~i;6.

These relations indicate that T,+ increases the number of particles in the system by
four and behaves like a scalar under SU(4). These properties of Tai become more
J. A. Evans et al. 1 Pairing correlations 83

TABLE 2
The classification according to pairing seniority (II,), S, T and N of the 1386 states spanning (6000)

01 s T N

0 0 6 12
5 10, 14
4 8, 12, 16
3 6, 10, 14, 18
2 4, 8, 12, 16, 20
1 2, 6, 10, 14, 18, 22
0 0,4,8,12,16,20,24

2 I 12
10, 14
8, 12, 16
6, 10, 14, 18
4, 8, 12, 16, 20
2, 6, 10, 14, 18, 22

0,2 12
IO,14
8, 12, 16
6, 10, 14, 18
4, 8, 12, 16, 20

l,3 12
10,14
8, 12, 16
6, 10, 14, 18

8 0,&J 2 12
1 10, 14
0 8, 12, 16

10 1, 3, 5 1 12
0 10, 14

12 0,2,4,6 0 12

The table may be used to deduce “deuteron” seniority (II,,) by interchanging S and T.

explicit when it is expressed in terms of the operators in (2a) giving

(13)

It is not difficult to show that the four-particle state created by applying T,+ to the
vacuum has unit norm.
The matrix elements of T,'in the basis IN, S, T,A,)may be calculated by writing
it in terms of the Casimir invariants of the orthogonal groups defined in (3). Thus

(14)
84 J. A. Evans et al. 1 Pairing correlations

where

i: (J,,)”
c, = p>q= 1

is the Casimir invariant of O(n). As C,, C, and Js7 are diagonal it follows
immediately that

Moreover, as C, is diagonal in the Gel’fand basis with eigenvalue &(A, + 5), the
matrix element in (15) is easily calculated from the eigenvectors of the two states
involved.

3. Results and discussion

All the results presented and discussed below are obtained for a configuration
space with degeneracy Q = 6. It is sufficiently large to allow the development of
collective (coherent) effects and yet the calculations can be performed without
heavy numerical work. In order to ensure that collective features emerge we con-
sider states in which the shell is about half full.
The analysis of the results is made separately for the energy spectra and for some
relevant matrix elements of two- and four-particle transfer reactions.

3.1. ENERGIES

As discussed in sect. 2 the hamiltonian (1) can be solved analytically within three
possible coupling schemes that correspond to the O(5) classifications for S(x = - 1)
or T(x = 1) and to tht: SU(4) Wigner supermultiplet (x = 0). We diagonalise energy
matrices constructed as explained in subsect. 2.2 and first focus our attention on
the spectra obtained for values of x corresponding to small deviations from those
coupling schemes.
In’ fig. 11we show the energy spectrum for x = -0.8 and twelve particles (half-
full shell).
We notice that this spectrum resembles, for a given S, a vibrational pattern.
There appear isospin multiplets each one labelled by the “isoscalar” seniority Q,.
For instance, for S = 0, we observe the ground-state T = 0 singlet (v,, = 0), the
T = 0,2 doublet (D,, = 4), the T = 0,2,4 triplet (uO = 8) and the T = 0,2,4,6
quadruplet (uO = 12). These multiplets can be interpreted respectively as zero-,
two-, four- and six-phonon states (i.e., u0 = 2ag, where ng is the number of phonons).
A rough estimate of their energies relative to the ground state can be obtained from
formula (lo),
E(v, = 2n,, S = 0) = $v,(4Q+6)-$v; = 252[n,(l+3/2Q)-n;/2Q]. (16)
J. A. Evans et al. / Pairing correlations 85

E-
60.- 6
/f
t
7
-2
0
40.- 4
---z

20.-

__--- -0
__--
___--
o” 1 2 3 4 5 6

Fig. 1. Energy levels for the system with N = 12 particles (half-full shell) and x = -0.8. The value
of T is indicated at the right of each level. Levels with the same S are displaced to the right and grouped
in columns. The S-value is indicated in the line below. The dashed line joins all the levels belonging to
the (even) yrast line and having a,, x 0. The remaining seniorities that are associated alternatively to odd-
and even-S bands are indicated above.

The departures from an equally spaced pattern, linear in ylg, are of order 52-i and
are due to both the Pauli exclusion principle and small dynamical effects that are
neglected in the lowest order collective treatment. Within this picture the one-
phonon state is identified with the lowest (S, T) = (0, 1) state in the system of ten
particles. We find that its energy turns out to be about half the energy of the T =
0,2doublet, thus supporting the vibrational interpretation. For a given Ta rotational
pattern in S can be observed in several parallel bands that approximately follow an
S(S+ 1) rule. Each can be labelled by the seniority v,, that becomes a goodquantum
number in the limit of x = - 1.
Within the present model an interchange of x by x and T by S leaves the result
-

unaltered. Hence the spin rotational pattern observed for x z - 1 appears again
for the isospin when x z 1.
The spectrum obtained when x = 0.2 is shown in fig. 2. In the vicinity of x = 0
the effects of the SU(4) symmetry are noticeable. The levels are grouped into sets
that can be labelled by & and follow approximately the &(A, +4) law of eq. (11).
From the point of view of the spin or isospin degrees of freedom, this region can
be regarded as phase transitional and consequently neither isoscalar nor isovector
coherent modes are present individually.
The progression from one limit to the other can be observed in fig. 3 where the
excitation energies of S = 0 states for twelve particles are plotted as a function of X.
86 J. A. Evans et al. 1 Pairing correlations

s= 0 1 2 3 4 5 6
T T T T T
E’ A 0'
c -0 -1 O_ [6E
-2
Z
--; -2 ,,’
-5 -_; : 1’ :l
30.- - 4
6 /’
/’
/’
/’
I’
//’
20
20.- -0 -1 ,’
-0
-2 -1 -2 /’ ,x [441

-4 -3 ,/
, I
/’
,’
10 - -d/b
-0 /’
- , I’ [221
A’
I’
c /’
-2
t
/’

0. /’ 0 [ool

Fig. 2. Energy levels for the system with N = 12 and x = 0.2. The conventions are the same as in fig. I,
The values of S are indicated in the line above. The W(4) label for each set of levels is shown. The dashed
line has the same meaning as in fig. 1.

T=2 --)_
J=O

-1. 0. Xl.
Fig. 3. Excitation energies of the S = 0 states for N = 12 and all possible values of T plotted against x.
J. A. Evans et al., / Pairing correlations 87

The pairing vibrational pattern holds reasonably well until x z 0. For x z 0,


isospin T(T + 1) bands start to develop, being completely established at x = 1.
Correspondingly a “deuteron” vibrational scheme emerges. Incidentally, we notice
also in this figure that for x z 0 both pairing modes are blurred.
In order to illustrate the changes from x = - 1 to x = 1, we study the structure
of the (S, T) = (0, 0) ground-state wave function as expanded in the basis with
good seniority u,,, i.e. the eigenfunctions of H for x = - 1. We notice (fig. 4) that

-.5 0. .5 X

Fig. 4. Absolute value of the components of the (S, 7) = (0, 0) ground-state wave function for N = 12
in the basis of good isoscalar seniority oO.

for the half-full shell the ground state is almost pure (2 80 %) ZJ~= 0 until values
of x close to x = 0. The overlap between the ground states for the two limiting
values of x is less than 1 ‘A. This result shows the quite different structure of the
deuteron and pair condensates. The phase transition (-0.2 5 x 5 0.2) is seen in
this graph as a region of rapidly changing components in the ground state.

3.2. TWO-PARTICLE TRANSFER

The fact that the hamiltonian given in (1) is constructed with the two-particle
operators P+ and Df makes pertinent the study of the transfer reactions that ex-
change a,pair of nucleons carrying zero orbital angular momentum and (S, T) =
(0, 1) or (1,0X
The amplitudes corresponding to the isoscalar pair transfer operator D+ have
already been obtained as ingredients of the calculation of the energy matrices. Thus
we can evaluate the matrix elements of Df for an arbitrary x once the eigenfunctions
of H have been found. Owing to the S-T symmetry of the model we need calculate
the matrix elements of only one of the two operators P+ and D+. The results for the
other are obtained by changing the sign of x and interchanging S and T. We choose
88 J. A. Evans et al. / Pairing correlations

_ypj--q

Fig. 5. Square of the two-nucleon T = 1 transfer matrix element on an N = 10 target populating states
in the N = 12 residual system, as a function of x. The full (dashed) lines correspond to transitions connect-
ing the lowest S = 0 (S = 2) states. Transitions under consideration are indicated in the level scheme
above.

to analyse the T = 1 pair transfer (P’) amplitudes because their behaviour can be
compared with previous studies that involve only the T = 1 pairing channel as a
residual interaction.
In fig. 5 we plot some of the relevant T = 1 spectroscopic amplitudes relating
systems with ten and twelve particles. We can observe two different types of variation
with X. The transitions labelled by A and C tend to a finite value in the pure pairing
limit (X = 1) while those labelled by B vanish in the same limit.
The first type of transition is enhanced for x = 1 due to the onset of a superfluid
phase which saturates at a value that depends on the number of the active pairs in
the condensate. In this sense the S = 0 (ground) states have more isovector pairs
than those with S = 2. This fact is reflected in the larger increase of the intensities
of the transition labelled A with respect to C for the (pairing) rotational extreme.
The second type of transition (B) behaves like those populating the two-phonon
pairing vibration in a pure pairing model. They tend to be enhanced in the phase
transitional region and vanish in the superfluid limit. This is due to the selection
rule imposed by the different isovector seniority u1 of the initial (N = 10) and final
(N = 12) states.
J. A. Evans et al. / Pairing correlations 89

In the pure isoscalar limit (X z - 1, weak pairing force) the intensities of the
transitions A, B and C have essentially the same value and all three can therefore
be interpreted as transfers involving the exchange of one phonon, thus supporting
the vibrational picture indicated by the energy spectra.

3.3. FOUR-PARTICLE TRANSFER

We now study the behaviour, as a function of x, of the amplitudes for the transfer
of four nucleons with the structure of an a-like particle (tig. 6). They are given by

(ST) - (00)
(00) -
- (00)
N=8 N=12 1
t<N+4,0,0iTa+lN,O,O>I
(10)-_- (10)
N=lO (C) N=14

Fig. 6. Absolute value of the a-transfer matrix element on N = 8 and N = 10 targets as a function of X.
Transitions under consideration are indicated in the level scheme above.

the matrix elements

(Nf4, S, TIT: IN, S, T>, (17)


which are easily calculated once the eigenfunctions of H have been found (see
subsect. 2.4).
The three amplitudes plotted are representative of the results obtained. Those
corresponding to transitions among states with T = S are even functions of x as
a consequence of the spin-isospin symmetry of the model.
For equal strength of the two forces (x = 0) a special situation arises. This is due
to the fact that the hamiltonian is diagonal in SU(4) and the transition operator
90 J. A. Evans et al. / Pairing correlations

T: is a scalar under that group [i.e. it commutes with all the infinitesimal operators
of SU(4)]. By considering the commutator [H(x = 0), Tz] we see that the role
played in this case by the a-transfer operator with respect to the hamiltonian is
similar to that played by the T = 1 two-particle operator with respect to the pairing
force [O(5)] or that of the quadrupole transition in regard to the Elliott quadrupole
interaction (SU(3)) [ref. ‘*)I. Within a collective picture the ;1, quantum number
may be thought as some kind of “intrinsic” structure label similar to the seniority
for the pairing model 6, or the (1, ,LL)quantum numbers for the SU(3) model. The
different behaviour of transition A (or C) and transition B in fig. 6 is due to the I,
selection rule and in this sense can be likened to the difference between transitions
AandBoffig.5inthelimitx= 1.
In order to understand the enhancement for x = 0 we need to estimate an upper
limit of the matrix element (15) in a situation similar to the one considered (N =
2Q-4). Such a value can be calculated if the effects of the exclusion principle are
neglected and [T,, T:] z 1. Then we can write the wave function of the ground
state of a system with N particles where N is a multiple of 4 as

IN) = {($V)l) -+(T,‘)N’410), (18)


and consequently, for N = 2G--4
(N+4, S, TjT,+IN, S, T) = ,/@. (19)

(S,T) _-_- - S=l transfers (D+)


ol-transfers
----- ......c..,... T.1 transfers(P+)

N:8 N.10 N.12

Fig. 7. Scheme of the two- and four-nucleon transfers connecting the ground state of the N = 8 system
with the lowest (S, 7) = (0,O) states of N = 12 particles. Strong matrix elements are indicated with a
heavier line. The values assigned to each transition are the squares of the corresponding matrix elements
for .x = - 0.8. The amplitudes for populating the states A and B following the different possible routes
are displayed in the inset.
J. A, Evans et al. / Pairing correlations 91

Our result for L? = 6 (curve ,A) is already 92 % of this upper limit and must get
closer to it for larger degeneracies where the exclusion principle effects are less
important.
We come now to the question of the relationship between the a-transfer and the
pairing degree of freedom 9-11). Within a pure isovector pairing model the CI-
transfer matrix element factorises into a product of two matrix elements of T = 1
pairs. Based upon this idea it has been suggested that the two-particle transfer
features associated with the pairing force would also be found in the four-body
transfer cross sections. The results shown in fig. 6 indicate, however, that other
channels of the residual force may invalidate to some degree this presumption.
In order to clarify this we show in fig. 7 a scheme of the transitions linking the
lowest (S, T) = (0, 0)states of eight and twelve particles. We choose again x =
-0.8, a situation in which the various excitations of the many-body systems can
be pictured as a deuteron condensate superimposed on a few, almost non-inter-
acting, T = 1 pairs, that produce a pairing vibrational pattern. The factorisation
mentioned above implies neglect of the first term in eq. (13). Introduction of a
complete set of N = 10 states having (S, T) = (0,1)between the P+ operators

3.0 I I I 6.0 ~
(N,S,T)-4N;S:T') +

w"

2 :
E 1.0 - : - 2.0 XJ
:
/ ________-----__-- 2
/'
I' .L,yO,O) -410,0,11 ;;I
$ ,.-' /
I' - 1.0 5
< _____..---
/'
_r-/ z
_______------
i;
0.0 I I I 6.0 w
-1.0 -0.5 0.0 0.5 1.0

X
Fig. 8. Squares of the a-transfer matrix elements populating the ground state of the N = 12 system
(scale at left) as a function of x. The figure also shows the square of the T = 1 two-nucleon transfer matrix
elements (scale at right) into which the a-transfer amplitude factorises approximately in the pairing extreme
x = 1. The labels on each curve are the (NST) of the target and residual states.
92 J. A. Evans et al. / Pairing correlations

leads to

(12;0,01T,+l8; o,o> = 4$C(12,0, OIIPfljlO, 0, 1, v)(lO; 0, 1, vlIP+/l8;0,0),

where v is a label which distinguishes the different intermediate states.


In the pairing extreme (X z l), the seniority (vl) selection rule reduces the sum
in eq. (20) to a single term, and in this limit isoscalar pair transfer is quite un-
important. Thus eq. (20) gives a reasonably good value for the cc-transfer amplitude.
In the case exhibited in fig. 7, however, the situation is quite different. Here eq. (20)
predicts roughly equal population for the ground and first excited (two pairing
phonon) states. This prediction coincides with the result of ref. 9, and with the
related assumption of ref. IL). When isoscalar pair transfer is included, it is seen to
be highly selective, taking full advantage of the existence of an isoscalar pair con-
densate and drastically enhancing the ground-to-ground transition. For this reason
the population of the ground state of N = 12 by a-transfer is considerably greater
than that of the excited pairing vibrational state.
The failure of eq. (20) to account for the a-transfer amplitude, except in the vicinity
of x = 1, may be appreciated by comparing the curves in fig. 8.

4. Conclusions

The aim of this paper has been to study the competition between the effects
induced by the isovector and isoscalar components of the residual nuclear force.
We find that the group O(8) permits exact treatment of a “model” hamiltonian
which retains most of the features of the actual residual interaction. The set of
generators of O(8) can be related to all the operators that are physically relevant
to the model. The labels for the O(8) basis give rise to selection rules for the transfer
operators that reduce to those of the well-known O(5) classification scheme for the
limits when only one (S, r) channel is active.
The relative variation of the strength of the “deuteron” term in the hamiltonian
produces a continuous change from a pairing vibrational to a superfluid (rotational)
situation. In this sense it may be compared to the effect of a single-particle splitting
in the pure pairing model. The plots displaying the behaviour of energies and two-
particle transition matrix elements (figs. 3 and 5) are in fact very similar to the
corresponding curves in the case of an exact diagonalisation of a T = 1 pairing
interaction in two levels [figs. 1 and 2 of ref. ‘j)].
The above similarity of the effects of a strong isoscalar component of the force
with respect to the isovector (pairing) channel and those produced by the single-
particle splitting can be understood in the light of the following considerations.
For x = - 1 only the isoscalar channel of the interaction is active bringing about
a condensate of (S, T) = (1,O) bosons (D-bosons) as the ground state of the system.
J. A. Evans et al. / Pairing correlations 93

Hence the enhancement of deuteron transfer is interpreted as due to the large number
of bosons 17) and yields a factor proportional to the number of bosons of the con-
densate. This ground-state band corresponds to states with spin seniority (zI,,)
zero. The excited seniority-two states can be thought of as a system where one D-
boson has been replaced by an isovector (0, 1) pair (P-boson) and similarly the
spin seniority four, six, etc. states correspond to the replacement of two, three, etc.
D-bosons by P-bosons. Taking into account the symmetry restriction that an even
(odd) number of P-bosons can only couple to even (odd) isospin Titis possible to
obtain a qualitative understanding of the spectra for x = - 1 and consequently
the similarity with the effects of the single-particle splitting.
The above parallelism between the effects of the isovector force in the presence
of the isoscalar channel or acting alone in a two-level model cannot be extended1
however to the analysis of four-body transfer matrix elements. In a model in which
only one (S, T) component of the two-body force is active the maximum values of
the cc-transfer cross sections are found at their rotational limits. In our calculation
instead the largest value is attained when both components of the force have equal
strengths. The reason for this stems from the fact that the pertinent transfer operator
is a scalar under the symmetry of the SU(4) subgroup of O(8), which is dominant in
this case. We can then conclude that the a-transfer process takes full advantage of
both the isoscalar and isovector terms of the residual interaction and is therefore
not specific to any single channel of the nuclear force. In this sense it may be mis-
leading to extend to a-transfer the results obtained for an operator specific to one
given component of the two-body residual interaction. We also have found that
the maximum cl-transfer is very close to the saturation value given by the transition
between a-condensates.

One of the authors (J.A.E.) wishes to thank the Comision National de Energia
Atomica for its kind hospitality and for the opportunity to participate in this work.

References

1) A. De Shaht and M. Goldhaber, Phys. Rev. 93 (1953) 1211


2) P. Federman and S. Pittel, Phys. Rev. C20 (1979) 820
3) K. T. Hecht, Phys. Rev. 139 (1965) B794
4) J. C. Parikh, Nucl. Phys. 63 (1965) 214
5) J. P. Elliott and J. A. Evans, Phys. Lett. 31B (1970) 157
6) G. G. Dussel, E. E. Maqueda and R. P. J. Perazzo, Nucl. Phys. A153 (1970) 469
7) D. R. B&s, E. E. Maqueda and R. P. J. Perazzo, Nucl. Phys. Al99 (1973) 193
8) B. H. Flowers and S. Szpikowski, Proc. Phys. Sot. 84 (1964) 673
9) 0. Dragdn, G. G. Dussel, E. E. Maqueda and R. P. J. Perazzo, Nucl. Phys. Al67 (1971) 529
10) F. D. Becchetti, L. T. Chua, J. Janecke and A. M. Vander Molen, Phys. Rev. Lett. 34 (1975) 225
11) R. A. Broglia, L. Ferreira, P. D. Kunz, H. Sofia and A. Vitturi, Phys. Lett. 79A (1978) 3.51 ;
A. Vitturi, L. Ferreira, P. D. Kunz, H. M. Sofia, P. F. Bortignon and R. A. Broglia Nucl. Phys. A340
(1980) 183
94 J. A. Evans et al. 1 Pairing correlations

12) M. K. F. Wong, J. Math. Phys. 8 (1967) 1899


13) S. Okubo, J. Math. Phys. 18 (1977) 2382
14) J. P. Elliott, Rendiconti della Scuola Internazionale di Fisica “E. Fermi”, XXXVI Corso, p. 128
15) S. C. Pang, Nucl. Phys. A128 (1969) 497
16) M. Hammermesh, Group theory (Addison-Wesley, Reading Mass., (1964) ch. 11
17) D. R. Bes, G. G. Dussel, E. E. Maqueda, and R. P. J. Perazzo, Nucl. Phys. Al27 (1973) 93
18) J. P. Elliott, Proc. Roy. Sot. A245 (1958) 128; 562

You might also like