perspec’ve)of)quantum decoherence)fromholography) and
TRANSCRIPT
![Page 1: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/1.jpg)
Perspec've of Quantum Decoherence from Holography
and Majorana Modes
Feng-‐Li Lin(NTNU &NCTS)
Based on 1. arXiv:1309.5855(JHEP 1401 (2014) 170) with Shih-‐Hao Ho(NCTS), Wei-‐Li(AEI-‐>ITP) and Bo Ning(NTNU-‐>SCU). 2. arXiv:1406.6249(NJP 16 (2014) 113062) with Shi-‐Hao Ho(NCTS), Sung-‐Po Chao(NCTS-‐>ASIOP) & Chung-‐Hsien Chou(NCKU) 3. on-‐going work with Pei-‐Hua Liu(NTNU) and Jen-‐Tsung Hsiang(Fudan U)
1
![Page 2: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/2.jpg)
Mo'va'on
1. Study the quantum decoherence of strongly coupled/correlated systems. E.g. Lu:nger liquid.
2. Bring a>en?on to the rela?on between Schwinger-‐Keldysh and Feynman-‐Vernon formalisms.
3. Iden?fy the ingredients for making the robust (topological) qubits against decoherence.
4. A novel way to tackle the decoherence of fermionic qubits.
2
![Page 3: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/3.jpg)
Outline
• I. Holographic decoherence
• II. Decoherence of fermionic
![Page 4: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/4.jpg)
Evolve reduced density matrix
• Total system = system (probe) + environment
• Ini?al state:
• Evolve total system:
• Evolve system:
• Master equa?on:
Non-‐unitary & Non-‐Markovian 4
![Page 5: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/5.jpg)
Formula'ons for decoherence
• There are different ways of formula?ng the framework for studying quantum decoherence.
1. A direct way is to solve the Heisenberg equa?ons of mo?on for both probe and environment simultaneously.
2. The most popular way is the so-‐called Feynman-‐Vernon formalism -‐-‐-‐ a path-‐integral approach to real-‐?me effec?ve field theory.
3. Evolve the reduced density matrix in the canonical formula?on. This is hard due to the complicated mixing normal-‐ordering among probe and environment operators.
4. For fermion, we find that we can use the Clifford algebra proper?es of Majorana modes to tackle the reduced dynamics in the canonical formula?on.
5
![Page 6: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/6.jpg)
Feynman-‐Vernon
• The task is how to evaluate the influence of the environment to the probe during evolu?on. • FV proposed a path-‐integral formula?on for evalua?ng the influence funcXonal. • The trick is to rewrite the forward and backward propagators in terms of path integral. • Then integra?ng out the environment dynamics.
6
![Page 7: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/7.jpg)
Keldysh contour
• This was done on a real-‐?me Keldysh contour:
7
![Page 8: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/8.jpg)
Feynman-‐Vernon formalism
• Using
• Then we have influence func?onal
• RDM:
• RDM propagator:
8
![Page 9: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/9.jpg)
FV vs Schwinger-‐Keldysh cf. Su, Chen, Yu, Zhou
• The FV influence func?onal can be compactly wri>en as:
for • This can be understood as the generaXng funcXonal for the real-‐Xme correlaXons of in SK formalism. • Here the probe acts as the external source to . • The probe loses its purity via a par?cular channel . • Using Kubo formula, one can then relate the transport coefficient to the retarded Green func?on in the context of linear response theory.
χO = − lim
ω→0limk→0
1ωImGO
R (ω ,k)
9
![Page 10: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/10.jpg)
SK formalism
• For Gaussian environment:
• By the causality principle, only 3 of the are independent. Perfoming Keldysh rota?on so that only & appear in the influence func?onal:
10
![Page 11: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/11.jpg)
I. Holographic Quantum Decoherence
11
![Page 12: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/12.jpg)
Holographic Influence Func'onal c.f. Maldacena, 1997; Gubser, Klebanov & Polyakov, 1998; WiWen 1998; Maldacena, 2003; Son & Herzog, 2003
Ø Influence func?onal = Genera?ng func?on of real-‐?me correla?on func?on.
Ø AdS/CFT duality: Effec?ve theory for a strongly coupled CFT is a co-‐dimensional 1 higher AdS gravity. Precisely,
Ø S?ll
12
![Page 13: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/13.jpg)
Holographic Green func'on
• GKP/W prescrip?on:
• Solve bulk EOMs:
• Linear response theory:
exp −Sbulk(on− shell ) φ[ ]( ) = exp φnn
(b)O∫ CFT
φ(t,x, z) φ (b)nn (t,
x)zΔnn + φ (b)n (t,x)zΔn ,Δn > Δnn .
source O(t,x) J(t,
x) response
O(ω ,k
) = −GR
O (ω ,k)J(ω ,k
)
χO = − lim
ω→0limk→0
1ωImGO
R (ω ,k)
For black hole, one needs to impose incoming b.c. to obtain retarded Green func?on.
13
![Page 14: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/14.jpg)
Holographic Environmental Kernels c.f. Son & Teaney, 2009; de Boer, Rangamani & Shigemori, 2009
1. The thermal kicks of the Hawking radia?on causes Brownian mo?on of the end-‐point.
The dissipa?on kernel for this holographic Brownian mo?on is super-‐Ohmic like:
2. C.f. the dissipa?on kernel for k=0 scalar mode in AdS5 is quasi-‐super-‐Ohmic:
GR(!) / (!2)
2��2[ln!2 � i⇡sgn(!)] for � � 2
14
![Page 15: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/15.jpg)
Environmental Kernels
• The above environmental kernels are of the Ohmic type spectra:
• This is the typical spectra for the helical Lu:nger liquid: Q=1 (Ohmic, Fermi-‐liquid), Q<1 (sub-‐Ohmic, a>arc?ve interac?on), Q>1 (super-‐Ohmic, repulsive interac?on).
• Non-‐Ohmic environmental kernel, such as the free spectra or the holographic ones, i.e., the holographic strange metal:
15 k<kF k>kF
![Page 16: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/16.jpg)
Holographic decoherence behaviors
16
![Page 17: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/17.jpg)
Characterize Decoherence (for con'nuous variables)
• Posi?vity of Wigner func?on (quantum version of phase space probability distribu?on func?on):
• For 2 Gaussian wave-‐packets (qubit of con?nuous variables):
17
![Page 18: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/18.jpg)
Evolve Wigner func'on
18
![Page 19: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/19.jpg)
Purity and Concurrence
n If the state of the qubits does not reduce to a pointer state, it means the decoherence is incomplete, and could be further purified. Then we can characterize the purity by
n For two-‐qubit state, we can characterize the quantum entanglement between the two qubits by concurrence:
n Zero concurrence implies no entanglement, but the state could be s?ll quantum.
are the square roots of eigenvalues, in decreasing order, of
![Page 20: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/20.jpg)
Generic Behaviors for W & S2
• Case 1. T=50(red), 30(green) and 0.1(blue)
• Case 2. Delta=4.5(green), 4(red) and 3.5(blue)
20
![Page 21: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/21.jpg)
Decoherence v.s. Local Quench
• Local quantum quench: A local excita?on is created suddenly and see how it relaxes back the equilibrium. • The thermaliza?on is characterized by a Gibbs state RDM. • If the local excita?on is a quantum state, then its purity will escape into the rest of the total system. This is the quantum decoherence. • The decoherence is characterized by a pointer state RDM.
21
![Page 22: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/22.jpg)
Decoherence v.s. local quench
• Local quantum quench (CFT + holographic calcula?ons):
• Quantum decoherence (Our fi:ngs):
22
c.f. Liu & Suh, 2014
![Page 23: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/23.jpg)
Scaling behaviors
• For Case 1, we find the decoherence ?me t_D is independent of temperature T at low T regime, and at high T regime, it behaves as
• As in CFT, there is only a ?me scale set by 1/T, which is approximately the thermaliza?on ?me scale. However, the ?me for the complete thermaliza?on for a con?nuous variable probe takes infinitely long.
• For Case 2, we consider T=0 case but with different conformal dimension Delta of operator O[x], we find
23
![Page 24: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/24.jpg)
II. Decoherence of Fermionic Qubits
24
![Page 25: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/25.jpg)
The generalized quadra'c fermion model
• The model Hamiltonian is for a (few) spinless fermionic qubit(s) interac?ng with the fermonic bath:
• Formally we can decompose the spinless fermions in to pairs of Majorana modes, and the Hamiltonian becomes
25
V = aO† +Oa†, {a, a†} = 1
O =X
!
[C(!)b! +D(!)b†!]
V = �1O1 + �2O2
�1 = a+ a†, �2 = i(a† � a)O1 = (O �O†)/2, O2 = (O +O†)/2i
H = Hp +HE + V a|1i = |0ia†|0i = |1i
![Page 26: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/26.jpg)
Issues about decoherence of fermionic qubits
• To adopt the path-‐integral Fenyman-‐Vernon method, we need to choose the coherent state basis with fermionic Grassmann eigenvalues.
• This was done in WM Zhang et al (2012, PRL) for the environment with Lorentzian spectra density.
• However, we found that the formalism fails to produce a posi?ve-‐definite RDM for large probe kine?c energy.
• We try to use alterna?ve way to evaluate RDM.
26
![Page 27: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/27.jpg)
27
Purity for RDM of a Dirac fermion probe in fermionic bath
![Page 28: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/28.jpg)
Canonical formula'on of RDM
• Instead of using Feynman-‐Vernon, we now use canonical formula?on in the interac?on picture. Then, the RDM takes the form:
• Even this is just a quadra?c model, it is quite difficult to evaluate RDM in a closed form due to the complica?on of mixed normal ordering for ``a” and ``O” operators.
28
⇢r(t) = e�iHpt(TrEU(t)⇢0U†(t))eiHpt
⇢0 = ⇢p ⌦ ⇢E , U(t) = Te�iR t d⌧VI(⌧)
VI(t) = ei(Hp+HE)t(aO† +Oa†)e�i(Hp+HE)t
![Page 29: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/29.jpg)
Dressing by the Majorana mode
• In terms of Majorana basis, we have so that the evolu?on operator is in the form Key Observa?on: Under Xme ordering and due to the dressing by Majorana modes, Q_a can be thought as the bosonic and mutually fermionic.
29
VI(t) = �1(t)O1(t) + �2(t)O2(t)
U(t) = Te�aQa(t), Qa(t) = i
Z
td⌧Qa(⌧)
T�1Q1(t)�1Q1(t) = �Z t Z t
[✓(⌧ � ⌧ 0)Q1(⌧)Q1(⌧0) + (⌧ $ ⌧ 0)]
T�1Q1(t)�2Q2(t) = T�2Q2(t)�1Q1(t)
= ��1�2R t R t
[✓(⌧ � ⌧ 0)Q1(⌧)Q2(⌧ 0)� ✓(⌧ 0 � ⌧)Q2(⌧ 0)Q!(⌧)]
![Page 30: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/30.jpg)
RDM in closed form
• By the above observa?on, we can eliminate gamma’s by Clifford algebra without worrying about the mixed normal ordering.
• In the mean-‐?me, just treat Q_a as the commu?ng operators as the bosonic/fermionic feature is implicitly hidden in the ?me ordering.
• In the end of the day, the RDM has the following form:
30
![Page 31: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/31.jpg)
31
Assume O_a are generalized free fields:
![Page 32: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/32.jpg)
32
S?ll in progress …
![Page 33: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/33.jpg)
33
III. Robust Topological Qubits
![Page 34: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/34.jpg)
Topological Qubits
• Consider a special case where the qubit is non-‐local, i.e., composed of 2 spa?ally separated Majorana modes. We call it topological qubit.
• We then couple this topological qubit to helical Lu:nger liquid with Ohmic type spectral density.
• As two Majorana modes are far separated, then <O_1 O_2>=0. Thus, the previous 4-‐point correlator becomes product of two 2-‐point ones.
• This kind of topological qubits can be realized as the edge modes of the Kitaev’s chain, 1d topological superconductor.
34
![Page 35: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/35.jpg)
35
![Page 36: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/36.jpg)
Special features
n There is no retarded Green func?on appearing in the final form of the reduced dynamics. This implies that the Majorana modes/Helical liquid are dissipa?onless, i.e., genera?ng no heat.
n The symmetric Green func?on appearing above is the Majorana-‐dressed one as discussed. It control the overall ?me dependence.
n Turn out that this ?me factor for Ohmic-‐like spectrum has a closed form, and has a cri?cal point at Q=1.
36
![Page 37: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/37.jpg)
Single topological qubit
1) Pure ini?al state:
2) For Z2-‐preserving interac?ons:
3) For Z2-‐breaking interac?ons:
4) The Z2 parity prevents from complete thermaliza?on into Gibbs state.
37
![Page 38: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/38.jpg)
Two Qubits: Special case for uniform environments
Before studying the reduced dynamics for more general initial states: Let us first consider a simple case: choose the initial state as , i.e.,
Fermionic:
Bosonic:
38
![Page 39: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/39.jpg)
Two T-‐qubits in fermonic environments
39
![Page 40: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/40.jpg)
C.f. Spin-‐Boson model ref. S.T. Wu PRA89p034301, also H.B. Liu et al, PRA87p052139
40
![Page 41: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/41.jpg)
Effec've gap-‐ness
n The quantum informa?on of the probe is carried away by the collec?ve excita?ons of the environment, which is specified by the spectral density.
n The Ohmic-‐like spectrum has no gap at low energy, and one would expect the complete decoherence.
n However, the super-‐Ohmic spectrum suppress more the low energy modes than the higher energy ones.
n Adding the topological nature of the Majorana modes, we see an effec?ve gap emerging for super-‐Ohmic cases.
41
![Page 42: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/42.jpg)
Non-‐uniform environments
n As the topological qubit is non-‐local, one can then couple its component Majorana modes to different environments. This is peculiar feature of the topological qubits.
n One can couple one of the Majorana modes to the sub-‐Ohmic environment and another one to the super-‐Ohmic one. Then, the decoherence pa>ern will depend on the ini?al states.
n Some peculiar pa>erns for the decoherece and concurrence can be obtained in such cases.
42
![Page 43: Perspec’ve)of)Quantum Decoherence)fromHolography) and](https://reader030.vdocuments.mx/reader030/viewer/2022012708/61a84e64a599304a614798f3/html5/thumbnails/43.jpg)
Summary
• The deocherence in holographic CFT environment is similar to the helical Lu:nger ones. It is interes?ng to consider other holographic environments.
• The trick of decomposing spinless fermion into 2 Majorana modes yields the closed form of RDM.
• Robust topological qubits are viable at least theore?cally.
43