Quantum Monte Carlo sign bounds topological Mott insulator and thermodynamic transitions in twisted bilayer graphene model Xu Zhang1Gaopei Pan2 3Bin-Bin Chen1Heqiu Li4Kai Sun5and Zi Yang Meng1

2025-04-29 0 0 1.8MB 13 页 10玖币
侵权投诉
Quantum Monte Carlo sign bounds, topological Mott insulator and thermodynamic
transitions in twisted bilayer graphene model
Xu Zhang,1Gaopei Pan,2, 3 Bin-Bin Chen,1Heqiu Li,4Kai Sun,5, and Zi Yang Meng1,
1Department of Physics and HKU-UCAS Joint Institute of Theoretical and Computational Physics,
The University of Hong Kong, Pokfulam Road, Hong Kong SAR, China
2Beijing National Laboratory for Condensed Matter Physics and Institute of Physics,
Chinese Academy of Sciences, Beijing 100190, China
3School of Physical Sciences, University of Chinese Academy of Sciences, Beijing 100049, China
4Department of Physics, University of Toronto, Toronto, Ontario M5S 1A7, Canada
5Department of Physics, University of Michigan, Ann Arbor, Michigan 48109, USA
(Dated: October 24, 2022)
We show that for magic-angle twisted bilayer graphene (TBG) away from charge neutrality, al-
though quantum Monte Carlo (QMC) simulations suffer from the sign problem, the computational
complexity is at most polynomial at certain integer fillings. For even integer fillings, this polyno-
mial complexity survives even if an extra inter-valley attractive interaction is introduced, on top of
Coulomb repulsions. This observation allows us to simulate magic-angle twisted bilayer graphene
and to obtain accurate phase diagram and dynamical properties. At the chiral limit and filling ν= 1,
the simulations reveals a thermodynamic transition separating metallic state and a C= 1 correlated
Chern insulator – topological Mott insulator (TMI) – and the pseudogap spectrum slightly above
the transition temperature. The ground state excitation spectra of the TMI exhibit a spin-valley
U(4) Goldstone mode and a time reversal restoring excitonic gap smaller than the single particle
gap. These results are qualitatively consistent with the recent experimental findings at zero-field
and ν= 1 filling in h-BN nonaligned TBG devices.
Introduction.— Magic-angle twisted bilayer graphene
(TBG) has attracted great attention in recent years,
as it hosts a variety of nontrivial phases beyond semi-
classical or band-theory description [1–41]. To theoreti-
cally characterize these flat bands and correlated quan-
tum phases, tight-binding [2, 5] and continuous (BM) [3]
models has been developed. In this study, we focus on
the continuous-model approach, which avoids the chal-
lenge to construct localized orbitals that preserves all the
symmetries [42–47]. By projecting long-range Coulomb
interactions onto the moir´e flat bands with proper quan-
tum metric, such projected Hamiltonian has been stud-
ied using mean-field approximations [20, 31, 48, 49]. At
certain limits, exact analytical solution has also been ob-
tained [50–54]. On the numerical side, the charge neu-
trality point has been studied using sign-problem-free
momentum-space quantum Monte Carlo (QMC) simula-
tions [55–57]. However, away from the charge neutrality
point, due to the arising of the sign problem, such simu-
lations have not yet been performed.
Although the sign problem often implies exponential
computational complexity, it is worthwhile to emphasize
that not all sign problems cause such severe damage.
Very recently, a much more mild type of sign problem
has been demonstrated, where the computation complex-
ity scales as a polynomial function of the system size,
known as polynomial sign problem [58, 59].
In this Letter, we study the sign problem in TBG flat
bands. Utilizing the sign bound theory [60], we prove
that TBG flat bands at even integer fillings, or arbitrary
integer fillings in the chiral limit, exhibit (at most) poly-
nomial sign problem. This observation allows us to utilize
QMC methods to study fillings away from charge neutral-
ity, away from the chiral limit, and/or in the presence of
extra attractive interactions on top of the Coulomb re-
pulsion (See Tab. I for details).
To demonstrate this new approach, we performed
large-scale QMC simulations to examine the chiral limit
at filing ν= 1 (we denote the fully empty/filled flat bands
as ν=4/+ 4 and charge neutrality as ν= 0). At
T= 0, this model can be solved exactly [51, 53, 56],
and the exact solution reveals that at T0, the sys-
tem is a correlated Chern insulator – a topological Mott
insulator (TMI) [61–66] – with Chern number C= 1.
Upon raising the temperature, the insulating state shall
“melt” into a metal and the time-reversal symmetry shall
be recovered. However, our knowledge about this finite-
temperature phase transition is very limited. Because
the sign problem is only polynomial, QMC simulations
become a highly efficient tool to study this system, from
which three different phases/states are observed (1) a
metal phase with time-reversal symmetry at high tem-
perature T > T ?, (2) a time-reversal invariant pseudo-
gap phase at intermediate temperature Tc< T < T ?
and (3) a low-temperature TMI phase at T < Tc. Here,
T?is a crossover temperature scale and Tcis the crit-
ical temperature of a second order phase transition, at
which the time-reversal symmetry is spontaneously bro-
ken. We further show that the TMI phase only breaks
the time reversal symmetry, while spins remain disor-
dered due to thermal excitations of gapless spin fluctua-
tions. This absence of spin order/polarization is in direct
contrast to quantum Hall states where electron spins are
polarized due to Zeeman splitting and thus spin fluctua-
arXiv:2210.11733v1 [cond-mat.str-el] 21 Oct 2022
2
tions are gapped. In the discussion section, we establish
the connection and consistency between our simulations
and recent experimental studies on TMI phases in h-BN
nonaligned TBG devices [41].
BM model and projected interaction.— We utilize the
BM model and project interactions between fermions to
the moir´e flat bands. The BM model Hamiltonian [3]
for the τvalley takes the following form Hτ
k,k0=
~vF(kK1)·σδk,k0V
V~vF(kK2)·σδk,k0
where K1and K2mark the two Dirac points
in the τvalley from layers 1 and 2 respectively.
V=U0δk,k0+U1δk,k0G1+U2δk,k0G1G2and
V=U
0δk,k0+U
1δk,k0+G1+U
2δk,k0+G1+G2
are the inter-layer tunnelings with matrixes
U0=u0u1
u1u0,U1=u0u1ei2π
3
u1ei2π
3u0and
U2=u0u1ei2π
3
u1ei2π
3u0where u0and u1are the
intra- and inter-sublattice inter-layer tunneling am-
plitudes. G1= (1/2,3/2),G2= (1,0) are the
reciprocal vectors of the moir´e Brillouin zone (mBZ),
and K1= (0,1/23)|G1,2|,K2= (0,1/23)|G1,2|.
For control parameters, we set (θ, ~vF/a0, u0, u1) =
(1.08,2.37745 eV,0 eV,0.11 eV) for the chiral limit, and
for non-chiral model, we set u0= 0.06 eV following
Refs. [50, 55, 57, 67]. Here, θis the twisting angle and
a0is the lattice constant of monolayer graphene.
For interactions, in addition to the Coulomb repulsion,
here we have the option to include one more interaction
term and the sign problem will still remain polynomial.
HI=1
2Ω X
q
(V1(q)δρ1,qδρ1,q+V2(q)δρ2,qδρ2,q)
δρ1,q=X
k,α,τ,s
(c
k,α,τ,sck+q,α,τ,s ν+ 4
8δq,0)
δρ2,q=X
k,α,s
(c
k,α,τ,sck+q,α,τ,s c
k,α,τ,sck+q,α,τ,s)
(1)
The first term in HI(V1>0) is the Coulomb interac-
tions, and the second term V20 introduces repulsive
interactions for fermions in the same valley and attrac-
tions between the two valleys, which can be introduced
as a phenomenology term describing inter-valley attrac-
tions [68–72]. At V2= 0, this model recovers the stan-
dard TBG model with Coulomb repulsions. When V2is
turned on, the inter-valley attraction favors inter-valley
pairing and could stabilize a superconducting ground
state. The normalization factor in HIis Ω = L23
2a2
M
with Lbeing the linear system size of the system. kand
qcover the whole momentum space, νis the filling factor
and α, τ, s represent layer/sublattice, valley, spin indices,
respectively. The mometum dependence for non-negative
V1and V2is unimportant to polynomial sign problem.
Here, for simplicity, we set V2=γV1with γbeing a
non-negative constant and for V1, we use Coulomb inter-
action screened by single gate in our simulation V1(q) =
e2
4πε Rd2r1
r1
r2+d2eiq·r=e2
2ε
1
|q|1e−|q|dwhere
d
2= 20 nm is the distance between graphene layer and
single gate, = 70is the dielectric constant. We then
project the interactions HIto the moir´e flat bands (See
SM [73]) and use the projected Hamiltonian to carry out
sign bounds analysis and QMC simulations.
Polynomial sign bounds.— In QMC simulations, the ex-
pectation value of a physical observable Ois measured as
hˆ
Oi=PlWlhˆ
Oil, where Wland hˆ
Oilare the weight and
the expectation value for the configuration l. Instead of
summing over all configurations, a QMC simulation sam-
ples the configuration space using the probability Wl. In
sign-problem-free QMC simulations, Wl0 for all land
an accurate expectation value can be obtained by only
sampling a small number of configurations – the impor-
tance sampling, and this number scales as a power-law
function of the system size. However, for many quantum
systems, Wlcan be negative or even complex, and thus
to obtain an accurate expectation value, it requires to
sample a large number of configurations, which usually
scales as an exponential function of the system size [74].
It is worthwhile to emphasize that the sign prob-
lem doesn’t always lead to an exponentially high com-
putational cost. To measure the severity of the sign
problem, here we use the average sign hsigni=
PlWl/Pl|Re(Wl)|, where |Re(Wl)|is the absolute
value of the real part of Wl. For physical partition func-
tion Z=PlWl=PlRe(Wl), this average sign is be-
tween 0 and 1, and hsigni= 1 means that the system
is sign problem free, while smaller hsignimeans sev-
erer sign problem. In a d-dimension quantum systems
that suffers from the sign problem, hsigni ∼ exp(βLd)
where β= 1/T the inverse temperature, indicating that
the number of configurations needed in QMC simulations
scales as an exponential function of the space-time vol-
ume. For polynomial sign problem, although hsigni<1
(i.e., the system does suffer from the sign problem),
1/hsigniis a polynomial function of the system size, and
thus the number of configurations needed only scales as
a power-law function of the system size.
Although the average sign can be easily measured in
QMC simulations, it usually doesn’t have a simple an-
alytic formula. To estimate the numerical cost to over-
come the sign problem, we utilize the sign bound hsignib
defined in Ref. [60]. As proved in Ref. [60], hsignibis
the lower bound of hsigni(i.e., hsignib6hsigni). Thus,
if the sign bound scale is a power-law function of the
system size, the sign problem is (at most) polynomial.
Remarkably, the low temperature sign bound in moir´e
flat bands can be easily calculated by counting ground
state degeneracy, which can be obtained using SU(4) and
SU(2) Young diagram as employed in Refs. [50, 51, 56]
3
TABLE I. Scaling of the sign bound hsignibat low tempera-
ture and large moir´e lattice size N=L2. A power-law func-
tion of Nindicates that the sign problem is (at most) polyno-
mial. 7indicates the sign bound decays to zero exponentially.
Filling(ν) Chiral(γ= 0) Non-chiral(γ= 0) Chiral(γ > 0)
0 1 1 1
±1N17 7
±2N2N1N2
±3N57 7
±4N8N4N4
(See SM for details [73]).
Details about this calculation are shown in the SM
and the conclusions are summarized in Tab. I. At charge
neutrality (ν= 0), moir´e flat bands with Coulomb inter-
actions (γ= 0) is known to be sign problem free [55, 56],
and thus the sign bound is 1. Here, we further prove
that adding inter-valley attractions (γ > 0) to the chiral
system doesn’t cause sign problem either (hsignib= 1).
Away from charge neutrality, sign problem arises, but it is
polynomial at certain integer fillings. For Coulomb repul-
sion (γ= 0), the sign problem is polynomial at any (even)
integer fillings at (away from) the chiral limit. When
inter-valley attractions are introduced (γ > 0), even in-
teger fillings at the chiral limit also have polynomial sign
bound.
To further verify the polynomial sign problem summa-
rized in Tab. I, we directly calculate hsigniand hsignib
in QMC simulations at various filling ν, and compare
them with the exact formula of the sign bound obtained
at integer fillings. As shown in Fig. 1, hsigniis always
larger or equal to hsignibas expected, and the sign bound
at integer filling indeed converges to the exact solution.
The peak in hsigniat integer (or even integer) fillings
indicates that the sign problem is less severe and QMC
simulations have a faster convergence at these fillings.
The location of these peaks are fully consistent with the
polynomial sign problem summarized in Tab. I. The only
exceptions are ν= 3 and 4 of Fig. 1(a), where the sign
problem is polynomial but the figure doesn’t exhibit vis-
ible peaks. This is because the power-law function here
has high powers N5and N8, which requires higher res-
olution (longer simulation time) to show clear distinction
from exponential functions.
Chiral limit at ν= 1 and γ= 0.— With the polynomial
sign bound obtained, here we discuss the QMC results as
a function of temperature for the chiral limit ν= 1 filling
case in this section. In the QMC simulation, we observe
a thermal phase transition with pseudogap spectrum and
spontaneous time reversal symmetry breaking, which is of
immediate relevance to the recent experimental finding of
correlated Chern insulators at zero-field and ν= 1 filling
in h-BN nonaligned TBG devices and its relatively high
Curie temperature of Tc4.5 K [41].
01234
0
0.2
0.4
0.6
0.8
1
01234
0
0.2
0.4
0.6
0.8
1
01234
0
0.2
0.4
0.6
0.8
1
01234
0
0.2
0.4
0.6
0.8
1
01234
0
0.2
0.4
0.6
0.8
1
01234
0
0.2
0.4
0.6
0.8
1
( )
a
( )
b
( )
c
( )
d
( )
e
( )
f
FIG. 1. hsigniversus filling ν(a,c,e) and hsignibversus filling
ν(b,d,f) at low temperature T= 1 meV. The fillings of ν < 0
are symmetric with respect to ν= 0 via the particle-hole
symmetry. (a-b) are the chiral limit γ= 0 cases, (c-d) are
the non-chiral γ= 0 cases where we take u0= 0.06 eV and
(e-f) are the chiral limit γ= 4 cases. (a,c,d) are the average
sign hsignifor L= 3(N= 9) and L= 4(N= 16) measured
from QMC for filling from ν= 0 to ν= 4. (b,d,f) are the
sign bounds hsignibfor L= 3(N= 9) and L= 4(N=
16) measured from QMC (solid line) and derived from exact
solution (ES) at the low temperature limit for filling ν=
1,2 (dash line values). The ES values in (b) are 1119/3278,
3630/111493 for L= 3 at ν= 1,2 and 945/4357, 183/14912
for L= 4 at ν= 1,2. The ES values are 1/10 for L= 3, 1/17
for L= 4 at ν= 2 in (d) and 100/2601 for L= 3, 71/5201
for L= 4 at ν= 2 in (f). (See SM for details [73])
At low temperature limit for ν= 1, exact solution
at T= 0 expects degenerate ground states with Chern
number C=±1, and ±3 [51, 53, 56]. In the large system
size limit N→ ∞, the number of ground states scales as
N7for C=±1 and N3for C=±3 [73]. Due to the
higher number of ground state degeneracy, thermal fluc-
tuations shall stabilize the C=±1 state as the thermal
equilibrium state at low temperature via the order by dis-
order mechanism [75]. This low-temperature state breaks
spontaneously the time-reversal symmetry, but this sym-
metry breaking process as a function of temperature is
unknown.
Our QMC simulation at finite temperature reveals this
process. To probe the time-reversal symmetry breaking,
we use the Chern band polarization as the order param-
摘要:

QuantumMonteCarlosignbounds,topologicalMottinsulatorandthermodynamictransitionsintwistedbilayergraphenemodelXuZhang,1GaopeiPan,2,3Bin-BinChen,1HeqiuLi,4KaiSun,5,andZiYangMeng1,y1DepartmentofPhysicsandHKU-UCASJointInstituteofTheoreticalandComputationalPhysics,TheUniversityofHongKong,PokfulamRoad,Hon...

展开>> 收起<<
Quantum Monte Carlo sign bounds topological Mott insulator and thermodynamic transitions in twisted bilayer graphene model Xu Zhang1Gaopei Pan2 3Bin-Bin Chen1Heqiu Li4Kai Sun5and Zi Yang Meng1.pdf

共13页,预览3页

还剩页未读, 继续阅读

声明:本站为文档C2C交易模式,即用户上传的文档直接被用户下载,本站只是中间服务平台,本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私,请立即通知玖贝云文库,我们立即给予删除!
分类:图书资源 价格:10玖币 属性:13 页 大小:1.8MB 格式:PDF 时间:2025-04-29

开通VIP享超值会员特权

  • 多端同步记录
  • 高速下载文档
  • 免费文档工具
  • 分享文档赚钱
  • 每日登录抽奖
  • 优质衍生服务
/ 13
客服
关注