Part 1: Why study ab initio Atomic Physics?

Part 2: SRG Methods Part 3: DHF + IMSRG

Exploring Open-Shell Atomic systems using Relativistic IMSRG

scheme

Nov 29, 2022

1 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

Table of Contents

1. Why study ab initio Atomic Physics?

1.1. ab initio survey of atomic species
1.2. Understanding nuclear phenomena with atomic calculations
1.3. Experiments need good theory!
2. SRG Methods
2.1. Flow equations and band-diagonalization
2.2. Normal ordered many-body Hamiltonian
3. DHF + IMSRG
3.1. DHF Hamiltonian in Second Quantization
3.2. Self-Consistent Field (SCF) Algorithm to compute DHF state
3.3. Recapitulation

2 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

Part 1: Why study ab initio Atomic Physics?

3 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

A map of computational hardness for atomic systems

Figure portrays the current hydrogen
1
helium
2

H He
landscape of computational 1.0079
lithium beryllium boron carbon nitrogen oxygen fluorine
4.0026
neon
3 4 5 6 7 8 9 10

load for different atomic Li

6.941
Be
9.0122
Computational difficulty increases B
10.811
C
12.011
N
14.007
O
15.999
F
18.998
Ne20.180
sodium magnesium aluminium silicon phosphorus sulfur chlorine argon

systems in their most stable 11

Na Mg
12

Al
13 14

Si
15

P S
16

Cl
17

Ar
18

22.990 24.305 26.982 28.086 30.974 32.065 35.453 39.948

states. Atoms coloured in potassium

19
calcium
20
scandium
21
titanium
22
vanadium
23
chromium
24
manganese
25
iron
26
cobalt
27
nickel
28
copper
29
zinc
30
gallium
31
germanium
32
arsenic
33
selenium
34
bromine
35
krypton
36

S D P
K Ca Sc Ti V Cr Mn Fe Co Ni Cu Zn Ga Ge As Se Br Kr
green are the ones where we 39.098
rubidium
37
40.078
strontium
38
44.956
yttrium
39
47.867
zirconium
40
50.942
niobium
41
51.996
molybdenum
42
54.938
technetium
43
55.845
ruthenium
44
58.933
rhodium
45
58.693
palladium
46
63.546
silver
47
65.38
cadmium
48
69.723
indium
49
72.64
tin
50
74.922
antimony
51
78.96
tellurium
52
79.904
iodine
53
83.798
xenon
54

already have a good Rb Sr

85.468 87.62
Y
88.906
Zr Nb Mo Tc Ru Rh Pd Ag Cd In
91.224 92.906 95.96 [98] 101.07 102.91 106.42 107.87 112.41 114.82
Sn Sb Te
118.71 121.76 127.60
I
126.90
Xe
131.29
caesium barium hafnium tantalum tungsten rhenium osmium iridium platinum gold mercury thallium lead bismuth polonium astatine radon
55 56 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86
understanding of their Cs Ba
132.91 137.33
Hf Ta
178.49 180.95
W Re Os
183.84 186.21 190.23
Ir
192.22
Pt Au Hg Tl
195.08 196.97 200.59 204.38
Pb Bi Po At Rn
207.2 208.98 [209] [210] [222]

structure through methods francium

87
radium
88
rutherfordium
104
dubnium
105
seaborgium
106
bohrium
107
hassium
108
meitnerium
109
darmstadtium roentgenium
110 111

Fr Ra Rf Db Sg Bh Hs Mt Ds Rg
like Coupled Cluster (CC) [223] [226] [261] [262] [266] [264] [277] [268] [271] [272]

which work well with closed lanthanum

57
cerium
58
praseodymium neodymium
59 60
promethium
61
samarium
62
europium
63
gadolinium
64
terbium
65
dysprosium
66
holmium
67
erbium
68
thulium
69
ytterbium
70
lutetium
71

F
La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu
shell and near-closed shell 138.91
actinium
89
140.12
thorium
90
140.91
protactinium
91
144.24
uranium
92
[145]
neptunium
93
150.36
plutonium
94
151.96
americium
95
157.25
curium
96
158.93
berkelium
97
162.50
californium
98
164.93
einsteinium
99
167.26
fermium
100
168.93
mendelevium
101
173.05
nobelium
102
174.97
lawrencium
103

Ac Th Pa U Np Pu Am Cm Bk Cf Es Fm Md No Lr
systems. [227] 232.04 231.04 238.03 [237] [244] [243] [247] [247] [251] [252] [257] [258] [259] [262]

4 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

Isotope Shift

* As we move down the period, atomic

sizes increase. Similarly, the size of the
nucleus also increases. This creates
possibility of multiple stable isotopes
of a single element.
* Different isotopes have slightly
different atomic spectra due to change
in the coulombic interaction due to
different nucleon number.
* Figure from [1].

5 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

EXPERIMENTS NEED GOOD THEORY!

6 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

What’s so hard about it?

* Electrons! Lot’s of them! Each interacting with the

nucleus and the other electron.
* Relativity! We are no longer trying to solve
Schrödinger equation (differential equation), but we
want to solve the Dirac equation (matrix
differential equation). Much much harder to solve!
* Nucleus is NOT point-like for larger atoms.
X N X
H= hi + gij (1)
i=1 i<j

7 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

8 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

Part 2: SRG Methods

9 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

SRG Flow Equations

* Many-body Hamiltonian is
band-diagonalized through
operation of unitaries. The
figure has been extracted
from [2].
*

H = Hd + Hod
lim H = Hd (2)
s→∞

*
d
H(s) = [η(s), H(s)] (3)
ds
10 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

[Technical] Choice of Generator and Magnus Expansion

Choice of generator η(s) is technically important but unenlightening. Any well defined
generator can solve the flow equation and allow IMSRG. However, some generators are
significantly more efficient than others depending on the application. Below are a few
historically popular generators.
* Wegner generator: η(s) = [Hd (s), Hod (s)]
* White generator: η Wh (s) = H∆(s)
od (s)

In order to solve the flow equation, we use the Magnus expansion. First we start by
considering an Ansatz H(s) = e Ω(s) and then expand the Hamiltonian in terms of Ω
∞
X
Ω = Ωn (4)
n=1
Z s Z s Z s1
1
Ω1 (s) = − ds1 η (s1 ) , Ω2 (s) = ds1 ds2 [η (s1 ) , η (s2 )] , . . . (5)
0 2 0 0
11 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

Many-Body Hamiltonian and Second quantization

The above equation is best solved in Second quantization. If you don’t know what that
means, here is a primer: Introduce a reference state |ψ⟩ which is generally (but not
always) the ground state of the desired Hamiltonian and define Fermionic ladder
operators that create the Hilbert space and satisfy the anti-commutation relations
{ai , aj } = {ai† , aj† } = 0, {ai† , aj } = δij (6)
A week later... write down the many-body Hamiltonian in Second quantization
1X
fij {ai† aj } + Γijkl {ai† aj† al ak }
X
H(s) = E0 (s) +
4
ij ijkl
1 X
+ Wijklmn {ai† aj† ak† al am an } + ... (7)
36
ijklmn

Ordering of ladder operators in the Hamiltonian is also important. We can use the
Wick’s theorem to normal order the operators. This is designated by the braces {}
12 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

Some comments on truncations and reference states

Second quantization is just another way of expressing QM, we haven’t solved our
problem of computational complexity. The infinite expasion of Hamiltonian given in
the previous slide is not solvable. Thus, we need to introduce truncations.
* Truncations have to be carefully done since we are dealing with operators. The
truncated series must be Normal ordered with respect to the reference state of the
system which we choose to be the DHF state |ψDHF ⟩ since it approximates the
ground state of the desired Hamiltonian. However, this does mean that we need
to compute |ψDHF ⟩ before we can perform IMSRG.
* While this seems tedious, there is a major computational advantage to normal
ordering the operators with respect to a reference state. We can induce a 3N
interactions without even computing the 3-body matrix elements Wijklmn . This is
not possible in most other RG methods and allows us to study much more
complex systems, including Open-Shell systems which have been practically
untouchable with almost all other methods.
13 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

It does work!

Figure: (a) Proficiency of IMSRG in nuclear physics [3] (b) Demonstration of principle... but
without relativity!

14 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

15 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

Part 3: DHF + IMSRG

16 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

Dirac-Hartree Fock Hamiltonian

Dirac Hartree Fock (DHF) Hamiltonian can be written as
H = H0 + V ; V = V0 + {V1 } + {V2 } (8)
where H0 represents the kinetic term of electrons, V0 , V1 , V2 are the zero-body,
one-body, two-body matrix elements respectively. They are given by
⃗ · p⃗ + (β − 1)mc 2 + Vcoulomb
H0 = c α
X 1X
V0 = − ⟨a| u |a⟩ + [⟨ab| u |ab⟩ − ⟨ba| u |ab⟩]
a
2
ab
X †
V1 = {ai aj } ⟨i| Vcoulomb |j⟩
ij
1X † †
V2 = {ai aj ak al } ⟨ij| Vcoulomb |kl⟩ (9)
2
ijkl

17 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

[Technical] Decomposing into radial & angular matrix elements

As noted above, in relativistic calculations,
  we are solving matrix differential equations.
P
Hence we pick a spinor ansatz that differentiates between the positive energy
iQ
components (P) and negative energy components (Q) of the Hamiltonian. Both P, Q
can be decomposed into radial and angular ansatzes. Radial functions are generally
chosen to be Laguerre functions encountered in solution of Hydrogen atom while the
angular components are the spherical harmonics. Making use of such a decomposition,
we can find the matrix elements of the Hamiltonian using the Wigner-Eckart theorem
m k j′
 
′ ′ ′
k
⟨γ, βκ, m| Sq (r ) γ , βκ , m = ⟨βκ| ∥ Sk ∥ βκ′
j q m′
⟨βκ| Sk βκ′ = Vk (γ, γ ′ , βκ, βκ′ ) ⟨βκ| ∥ Ck ∥ βκ′
Z ∞
Vk (γ, γ ′ , βκ, βκ′ ) = ∗
Rγ,βκ (r )Rγ ′ ,βκ′ (r )vk (r )dr (10)
0
18 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

Solution to Roothaan Equation and SCF Algorithm

The operators introduced above (H0 , Vi ) create the Fock matrix F which characterises
the interactions. The Fock matrix does not satisfy an eigenvalue equation but rather
the Roothaan equation
FC = ϵSC (11)
where S is the overlap matrix. Thus, the above equation needs to be solved using an
iterative self-consistent field algorithm to obtain the DHF state |ψDHF ⟩. Details of the
algorithm do not provide much insight into the problem and thus we refer the reader to
[4] for details of the algorithm.

19 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

Recapitulation and Outlook

Outline of the procedure:
hydrogen helium

* Find the MEs of the DHF H

1

1.0079
He
2

4.0026
lithium beryllium boron carbon nitrogen oxygen fluorine neon

Hamiltonian using Li
3

Be
4

Computational difficulty increases B

5

C
6

N
7

O
8

F
9

Ne
10

6.941 9.0122 10.811 12.011 14.007 15.999 18.998 20.180

Wigner-Eckart theorem sodium

11

Na Mg
magnesium
12
aluminium

Al
13
silicon
14

Si
phosphorus
15

P
sulfur

S
16
chlorine

Cl
17
argon

Ar
18

22.990 24.305 26.982 28.086 30.974 32.065 35.453 39.948

potassium calcium scandium titanium vanadium chromium manganese iron cobalt nickel copper zinc gallium germanium arsenic selenium bromine krypton

* Compute |ψDHF ⟩ using SCF 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

S D P
K
39.098
Ca Sc
40.078 44.956
Ti
47.867
V
50.942
Cr Mn Fe Co Ni Cu Zn Ga Ge As Se Br
51.996 54.938 55.845 58.933 58.693 63.546 65.38 69.723 72.64 74.922 78.96 79.904
Kr
83.798

algorithm rubidium
37

Rb Sr
strontium
38
yttrium
39

Y
zirconium
40
niobium
41
molybdenum
42

Zr Nb Mo Tc Ru Rh Pd Ag Cd In
technetium
43
ruthenium
44
rhodium
45
palladium
46
silver
47
cadmium
48
indium
49
tin
50

Sn Sb Te
antimony
51
tellurium
52
iodine
53

I
xenon

Xe
54

85.468 87.62 88.906 91.224 92.906 95.96 [98] 101.07 102.91 106.42 107.87 112.41 114.82 118.71 121.76 127.60 126.90 131.29
caesium barium hafnium tantalum tungsten rhenium osmium iridium platinum gold mercury thallium lead bismuth polonium astatine radon

* Use |ψDHF ⟩ as a reference 55

Cs Ba
132.91
56

137.33
72

Hf Ta
178.49
73

180.95
74

W Re Os
183.84
75

186.21
76

190.23
77

Ir
192.22
78

Pt Au Hg Tl
195.08
79

196.97
80

200.59
81

204.38
82

Pb Bi Po At Rn
207.2
83

208.98
84

[209]
85

[210]
86

[222]

state to perform IMSRG and

francium radium rutherfordium dubnium seaborgium bohrium hassium meitnerium darmstadtium roentgenium
87 88 104 105 106 107 108 109 110 111

Fr Ra
[223] [226]
Rf Db Sg Bh Hs Mt Ds Rg
[261] [262] [266] [264] [277] [268] [271] [272]

compute the correlation lanthanum cerium praseodymium neodymium promethium samarium europium gadolinium terbium dysprosium holmium erbium thulium ytterbium lutetium
57 58 59 60 61 62 63 64 65 66 67 68 69 70 71

F
energies between electrons. La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu
138.91 140.12 140.91 144.24 [145] 150.36 151.96 157.25 158.93 162.50 164.93 167.26 168.93 173.05 174.97
actinium thorium protactinium uranium neptunium plutonium americium curium berkelium californium einsteinium fermium mendelevium nobelium lawrencium
89 90 91 92 93 94 95 96 97 98 99 100 101 102 103

* Add those energies on top of Ac Th Pa

[227] 232.04 231.04
U
238.03
Np Pu Am Cm Bk Cf
[237] [244] [243] [247] [247] [251]
Es Fm Md No Lr
[252] [257] [258] [259] [262]

the DHF calculation.

20 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

References

Nakhate, S.G., Bhattacharyya, S., Mukund, S. et al. Isotope-selective laser

photoionization of tin in supersonic atomic beam. Appl. Phys. B 125, 23 (2019).
T. D. Morris and S. K. Bogner , ”The Magnus expansion and the in-medium
similarity renormalization group”, AIP Conference Proceedings 1619, 117-125
(2014).
Hergert H (2020) A Guided Tour of ab initio Nuclear Many-Body Theory. Front.
Phys. 8:379.
C. C. J. Roothaan, ”Self-Consistent Field Theory for Open Shells of Electronic
Systems”, Rev. Mod. Phys. 32, 179

21 / 22
Part 1: Why study ab initio Atomic Physics? Part 2: SRG Methods Part 3: DHF + IMSRG

Thank You!

22 / 22