Coupled-cluster

effective particle-hole interaction

Zhonghao Sun (孙中浩)

UTK/ORNL
 • Shell model coupled cluster
method.
• Multi-shell effective interaction.
Outline • Coupled-Cluster particle-hole
effective interaction .
• Preliminary calculations.
• Summary
Shell model coupled cluster method:
Close-shell(single reference, CC) decoupling:
T 1
• Similarity, H̄ = e HeT = H + [H, T ] + [[H, T ], T ] + ... = (HeT )C
2!
• non-Hermitian h 0 |e
T
HeT | 0i =E T̂ =
X
Ti
• Advantages and disadvantages in h ab... T
HeT |
i
ij... |e 0i =0
truncation(bi-orthogonal) X X
1
• Controllable errors Tˆ1 = ti ↵a† ↵i , Tˆ2 =
2
tab † †
ij ↵a ↵b ↵i ↵j
ia i<j,a<b

Shell model couple-cluster: ¯ =e

H̄ S
H̄eS

S = QSP <latexit sha1_base64="TVt3KL5BwuWb4UVFecQ3Q0wx/CM=">AAAB7HicbVDLSgMxFL1TX7W+qi7dBIvgqsz4QDdC0Y3LljptoR0kk2ba0EwyJBmhDP0GNy4UcesHufNvTNtZaOuBwOGcc8m9J0w408Z1v53Cyura+kZxs7S1vbO7V94/aGmZKkJ9IrlUnRBrypmgvmGG006iKI5DTtvh6G7qt5+o0kyKBzNOaBDjgWARI9hYyW/eNJr1x3LFrbozoGXi5aQCOWz+q9eXJI2pMIRjrbuem5ggw8owwumk1Es1TTAZ4QHtWipwTHWQzZadoBOr9FEklX3CoJn6eyLDsdbjOLTJGJuhXvSm4n9eNzXRdZAxkaSGCjL/KEo5MhJNL0d9pigxfGwJJorZXREZYoWJsf2UbAne4snLpHVW9c6rl42LSu02r6MIR3AMp+DBFdTgHurgAwEGz/AKb45wXpx352MeLTj5zCH8gfP5AxVujjc=</latexit>

• Similarity, X X X
a⌫ abµ⌫
S1 = S , S2 = S + Siab⌫
• non-Hermitian a⌫ abµ⌫ abi⌫
• disadvantages in truncation(bi-
<latexit sha1_base64="qkgSaxQ0mBXXtflC/tQ/4lEVl2g=">AAACPHicbZBLSwMxFIUzPmt9VV26CRZBUMpMVXQjFN24rNQ+oFOHTJq2oUlmyEMoQ3+YG3+EO1duXCji1rXpS7T1QuDc8+WS3BPGjCrtus/O3PzC4tJyaiW9ura+sZnZ2q6oyEhMyjhikayFSBFGBSlrqhmpxZIgHjJSDbtXA169J1LRSNzqXkwaHLUFbVGMtLWCTKkUeBe+MjxIkC9MH5buRuIIloL8hIQ+Nz900hxOIB10I2JFQINM1s25w4KzwhuLLBhXMcg8+c0IG06ExgwpVffcWDcSJDXFjPTTvlEkRriL2qRupUCcqEYyXL4P963ThK1I2iM0HLq/JxLElerx0N7kSHfUNBuY/7G60a3zRkJFbDQRePRQyzCoIzhIEjapJFiznhUIS2r/CnEHSYS1zTttQ/CmV54VlXzOO86d3pxkC5fjOFJgF+yBA+CBM1AA16AIygCDB/AC3sC78+i8Oh/O5+jqnDOe2QF/yvn6BhxWr/A=</latexit>

orthogonal)
• Controllable errors
• Open-shell
• Fails in multi-shell decoupling
• Convergency property(largest
overlap with the model space)
|Ψ0!

0ph(2p0h) 1p1h(3p1h) 2ph(4p2h)

Shell model coupled cluster Numerical implement
¯ = H̄ + [H̄ , S ] + 1 [[H̄ , S ] , S ] + ...
The second similarity transformed H̄ 2b 2b 2b 2b
2!
2b 2b 2b 2b 2b

Hamiltonian : 1
+[H̄2b , S2b ]3b + [[H̄2b , S2b ]2b , S2b ]3b + ...
2!
[H̄2b , S2b ]2b 1
+ [[H̄2b , S2b ]3b , S2b ]2b + ...
2!
1
+ [[H̄2b , S2b ]3b , S2b ]3b + ...
2!

• Better convergent, but

no solution to intruder
states,
• limit contribution to
[[H̄2b , S2b ]3b , S2b ]2b excited states

[H̄2b , S2b ]3b

A1 A3 A5 d1
B1

d2
• Improved energy
level even on leading
A2 A4 A6

order.
B2

• Heavy computational
d3
C1 C3 C5

load in iterative
B3

d4
C2 C4 C6
 E(MeV)

-28
-27
-26
-25
-24
-23
-22
-21
EOM-CC

0
+
2+
SMCC(2b)

0+
2+
SMCC(3b-diag)
6

2+

0+
He

SMCC(3b-od)

0
+
2+

FCI

0+
2+

VS-IMSRG
0+
2+
How we are doing:

E(MeV)
-30
-25
-20
-15

EOM-CCSD
0 +
1+

EOM-CCSDT1
0 +
1+

CCEI
1

0+
+
8

SMCC(3b-od)
He

0+
1+

SMCC(3b)
0+
1+

FCI
0+
1+

VS-IMSRG
0+
1+

E(MeV)
-45
-40
-35
-30
-25
-20

EOM-CCSD

EOM-CCSDT1

CCEI
8

SMCC(3b-od)
Li

SMCC(3b)

FCI

VS-IMSRG
Where we are

• With perturbative treatment of off-diagonal and diagonal induced three-body forces, we

get good agreement with NCSM
• No convergent problem in most good close shell nuclei
• Some special very sensitive nuclei, iterative three-body level decoupling is still
desirable(O26 resonant ) to get reasonable result
• The property of convergent is similar to most effective interaction method, it converge to
states that have largest overlap with the model space.
• Preliminary iterative three-body level decoupling do not solve the multi-shell problem.
No convergent or over-rotated. Near-divergent two-body part and three-body part do not
naively cancel.
Same conclusion from im-SRG Approach

From Ragnar’s talk

 good shell mode
Well defined fermi
nuclei, induced three-
surface, good shell
body force not
mode nuclei
neglectable

Good close shell

nuclei, complex excited
state(O16)

Not good shell nodel

nuclei
shell model coupled cluster particle and hole effective interaction

¯ =e
H̄ S
H̄eS S = Spp + Shh + Shp
<latexit sha1_base64="wmY8h03BDvbVCcRp7eJwdFe9hSQ=">AAACFHicbVDLSsNAFJ34rPUVdekmWAShUBIf6EYounFZqX1AG8JkOmmGTh7M3Igl5CPc+CtuXCji1oU7/8ZpG8S2Hhg495x7uXOPG3MmwTS/tYXFpeWV1cJacX1jc2tb39ltyigRhDZIxCPRdrGknIW0AQw4bceC4sDltOUOrkd+654KyaLwDoYxtQPcD5nHCAYlOXq5fll30i7QB0jjOMvKv5XvT1XKc/SSWTHHMOaJlZMSylFz9K9uLyJJQEMgHEvZscwY7BQLYITTrNhNJI0xGeA+7Sga4oBKOx0flRmHSukZXiTUC8EYq38nUhxIOQxc1Rlg8OWsNxL/8zoJeBd2ysI4ARqSySIv4QZExigho8cEJcCHimAimPqrQXwsMAGVY1GFYM2ePE+axxXrpHJ2e1qqXuVxFNA+OkBHyELnqIpuUA01EEGP6Bm9ojftSXvR3rWPSeuCls/soSlonz8HM6AX</latexit>

With hole-hole term in the generator, the Baker-Campbell-Hausdorff needs even

more terminate with operator rank truncation

The decoupling will provide

1. hh interaction to calculate particle remove system(A-1, A-2, A-3, A-4, ....), vs PR-CC
2. pp interaction for particle attached system( shell model interaction), vs PA-CC
3. hp interaction for charge exchange system.
4. hp for particle-hole shell model to calculate excited states of close shell nuclei
5. We need shell model code to treate particle and hold at same time, or re-normal
order to get pp only interaction
S3 contributes to ¯
H̄
<latexit sha1_base64="Qe4CTtWj4+VwPtb/i3O8Stcvb/M=">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</latexit>
2b

Three-body off-diagonal diagrams in SMCC

A1 A3 A5 d1
B1

d2
A2 A4 A6

B2

d3
C1 C3 C5

B3

d4
C2 C4 C6
 Preliminary result
5

-5
E(MeV)

-10
S(n) of Z=50
14
-15 Exp
1.8/2.0(EM)
delta394 12

E(MeV)
-20

10
-25
100 102 98 100
Sn Sn Cd In
8
51 52 53 54 55 56
neutron number N Exp
S(p) of N=50 1.8/2.0(Em)
6 delta394

Binding Energy respect to Sn100

4

E(MeV)
2

0
50 49 48 47 46
proton number, Z

Its not a good idea to dig a lot of holes on a close-shell nuclei

Preliminary result

102 104 106

3 Sn Sn 6+ Sn
6+ 4+
6+ 4+
2.5
4+
6+ 6+
2+ 2+
2 2+ 4+ 6+ 6+ 4+ 6+ 4+
6+ 4+ 4+
E(MeV)

1.5 4+ 2+
2+ 2+
2+ 2+
1 2+

0.5

0 0+ 0+ 0+ 0+ 0+ 0+ 0+ 0+ 0+
1.

de

1.

de

1.

de
ex
ex

ex
8/

8
lta

l
p
p

p
ta

ta
/2

/2
2.

39

3
.0

.0
0

94

94
(E

(
4

EM

EM
M
)

)
Preliminary result

3
98 100 102
Cd Cd Cd

2.5 6+
6+
4+ 6+ 6+ 6+
4+ 4+
4+ 6+
2
4+
6+
2+ 4+ 4+
E[MeV]

1.5 6+ 6+
2+ 2+ 4+
4+
2+
1 2+
2+ 2+ 2+
2+

0.5

0 0+ 0+ 0+ 0+ 0+ 0+ 0+ 0+ 0+
1.

1.

1.
ex

de

ex
de

de

ex
8

8
p

p
/2

/2

/2
l
lta

p
lta
ta
.0

.0

.0
39
39

39
(E

(E

(E
4
4

4
M

M
)

)
Preliminary result

101 103
100
In In In
11/2+
1.8 13/2+
1/2-
1/2-
1.6

1.4
13/2+
11/2+
1.2

11/2+
1 13/2+
E(MeV)

0.8

2+
0.6 2+ 1/2-
3+
1/2-
0.4
3+
5+ 2+
4+ 3+
0.2 6+ 5+
6+ 1/2-
4+ 4+
0 7+ 7+ (6/5+) 9/2+ 9/2+ 9/2+ 9/2+ 9/2+ 9/2+

-0.2
1.

1.

1.

de
de

de

ex

ex
ex
8/

8/

8/

lta
p

p
lta

lta
p
2.

2.

2.

39
39

39
0(

0(

0(

4
EM

EM

EM
4

4
)

)
 • Particle hole effective interaction
for shell model.
• Non-Hermitian shell model
code, computational tough.
Summary • Iterative three-body level
decoupling
Thanks