Article https://doi.org/10.

1038/s41467-024-52927-w

Universality of quantum phase transitions in

the integer and fractional quantum Hall
regimes
Received: 6 May 2024 Simrandeep Kaur1,8, Tanima Chanda 1,8, Kazi Rafsanjani Amin2,8, Divya Sahani1,
Kenji Watanabe 3, Takashi Taniguchi 4, Unmesh Ghorai5, Yuval Gefen 6,
Accepted: 24 September 2024
G. J. Sreejith7 & Aveek Bid 1

Check for updates Fractional quantum Hall (FQH) phases emerge due to strong electronic
1234567890():,;
1234567890():,;

interactions and are characterized by anyonic quasiparticles, each dis-

tinguished by unique topological parameters, fractional charge, and statistics.
In contrast, the integer quantum Hall (IQH) effects can be understood from the
band topology of non-interacting electrons. We report a surprising super-
universality of the critical behavior across all FQH and IQH transitions. Con-
trary to the anticipated state-dependent critical exponents, our findings reveal
the same critical scaling exponent κ = 0.41 ± 0.02 and localization length
exponent γ = 2.4 ± 0.2 for fractional and integer quantum Hall transitions.
From these, we extract the value of the dynamical exponent z ≈ 1. We have
achieved this in ultra-high mobility trilayer graphene devices with a metallic
screening layer close to the conduction channels. The observation of these
global critical exponents across various quantum Hall phase transitions was
masked in previous studies by significant sample-to-sample variation in the
measured values of κ in conventional semiconductor heterostructures, where
long-range correlated disorder dominates. We show that the robust scaling
exponents are valid in the limit of short-range disorder correlations.

The quantum Hall (QH) effect, observed in a two-dimensional electron longitudinal resistance Rxx. As the Fermi energy approaches Ec, the
gas subject to a perpendicular magnetic field, realizes multiple quan- localization length ξ characterizing the single-particle states diverges
tum phase transitions (QPT) between distinct insulating topological as ξ ~ ∣E−Ec∣−γ while the slowest time-scale diverges as τ ~ ξz ~ ∣E−Ec∣−zγ8,9.
states1. The magnetic field B quenches the electronic kinetic energy The exponent γ governs the critical divergence of the localization
into disorder-broadened discrete Landau energy levels (LL). All elec- length as the filling fraction or magnetic field approach the critical
tronic single-particle states are localized, barring those at a specific values and z governs the divergence of the coherence length with
critical energy Ec near the center of each LL, which are extended2–7. decreasing temperatures10. From the finite-size scaling theory10,11,
When the Fermi energy lies between the extended states of two suc-
cessive LLs, the system is in a distinct topological phase characterized
by a quantized value of Hall resistance Rxy and vanishingly small dRxy =dν ν = ν c / T 1=zγ ð1Þ

1
Department of Physics, Indian Institute of Science, Bangalore 560012, India. 2Department of Microtechnology and Nanoscience, Chalmers University of
Technology, 412 96 Gothenburg, Sweden. 3Research Center for Electronic and Optical Materials, National Institute for Materials Science, 1-1 Namiki, Tsukuba
305-0044, Japan. 4Research Center for Materials Nanoarchitectonics, National Institute for Materials Science, 1-1 Namiki, Tsukuba 305-0044, Japan.
5
Department of Theoretical Physics, Tata Institute of Fundamental Research, Homi Bhabha Road, Mumbai 400005, India. 6Department of Condensed Matter
Physics, Weizmann Institute of Science, Rehovot 76100, Israel. 7Indian Institute of Science Education and Research, Pune 411008, India. 8These authors
contributed equally: Simrandeep Kaur, Tanima Chanda, Kazi Rafsanjani Amin. e-mail: aveek@iisc.ac.in

Nature Communications | (2024)15:8535 1

Article https://doi.org/10.1038/s41467-024-52927-w

Here, ν = nh/eB, n is the areal charge-carrier density, h is the Planck expected picture of multiple plateau-to-plateau quantum phase tran-
constant, e is the electronic charge, and T is the temperature. One sitions, each with its own distinct critical properties, here we find that
additionally defines the scaling exponent κ = 1/zγ11–13 that governs this for all IQH and FQH plateau-to-plateau transitions (PT), κ = 0.41 ± 0.02,
temperature dependence of the slope of Rxy as well as the width of the γ ≈ 2.4 ± 0.2, and z ≈ 1, closely aligned with the predictions of the scal-
Rxx peak at the transition. The values of these three critical exponents ing theory of localization13. Given the distinct origins of the two phe-
(of which only two are independent) have been argued to be universal, nomena, this striking similarity of the critical exponents suggests a
with γ ≈ 2.3, κ ≈ 0.42, and z = 1 for all IQH transitions4,10,11,14,15. connection between the IQH and FQH effects that transcends the
Low temperatures and high magnetic fields enhance the effective composite fermion (CF) framework.
electron-electron interactions, producing a richer set of the fractional We estimate the values of κ near criticality (ν ≈ νc) using three dis-
quantum Hall (FQH) phases at rational filling fractions16. The question tinct approaches: (i) analyzing the critical divergence of dRxy/dν, (ii)
then arises: Can IQH and FQH phase transitions be analyzed using a probing the critical divergence of the inverse width of Rxx(T), and (iii) a
‘unified’ scaling framework17? While the IQH phases originate from the scaling analysis of Rxy near the critical point. The localization exponent γ
topology of the single particle electronic Chern bands18, the FQH phases is obtained deep in the tails of the localized regime from the dependence
are crucially underlain by strong electronic interactions. These are of Gxx on ν. A scaling analysis of Quantum Hall transitions for fractional
marked by distinct electronic correlations, topological order, ground and integer states provides a second, independent way to extract γ.
state degeneracy, and topological entanglement. The transition The realizations of these quantum phase transitions in graphene-
between FQH plateaus is driven by a proliferation of anyonic quasi- based systems are associated with a highly tunable set of parameters.
particles (characterized by quasiparticle statistics and fractional charge). These include the ability to alter electron density, which is typically
This picture may suggest that the critical behavior at the transitions unachievable in semiconductor heterostructures25, the capability to
depends on the specifics of the topological FQH states involved and is manage screening, and the option to induce band mixing by applying a
also different from the analogous transitions in the IQH regime. displacement field D. This flexibility helps us establish that weak Landau
Experimental investigations of scaling in the IQH regime have level mixing does not significantly affect these critical exponents.
reported κ varying between 0.16 ≤ κ ≤ 0.81 (Supplementary Informa- Graphene also provides a platform where the nature of disorder
tion, Supplementary Note 13). This wide variation has been attributed scattering can be controlled. This is because the electrical transport
to varying disorder correlation lengths with a universal critical beha- properties of high-mobility graphene devices are dominated by short-
vior seen only in samples with short-range disorder19,20. This lack of a range impurity scattering, while those of low-mobility graphene devi-
tight constraint on κ has hindered any claims of their universality. ces are controlled by both short-ranged and long-ranged scattering
Similar experimental investigations of scaling laws at transitions potentials26,27. Thus, high-mobility graphene devices represent a nat-
between FQH phases are scarce21–23. A recent experimental study on ural candidate to investigate the universality of scaling exponents. Our
extremely high-mobility 2D electron gas confined to GaAs quantum comparative study between graphene devices of varying mobility
wells found the value of κ in the FQH regime to be non-universal, this shows that as long as long-range impurity scattering can be sup-
observation being attributable to long-range disorder correlation23. pressed, the universality of scaling parameters persists, independent
Thus, despite over three decades of study, the fundamental question of the quantum Hall bulk phases involved.
of the values of the critical exponents across quantum Hall transitions
(integer and fractional) remains unsettled14,23,24. Results
This article reports the experimental observation of a surprising Standard dry transfer technique is used for the fabrication of dual
super universality in the scaling exponents for transitions between graphite-gated hexagonal-boron-nitride (hBN) encapsulated TLG
various IQH and FQH phases in trilayer graphene. We measure both the devices (Fig. 1a) (for details, see Supplementary Information, Supple-
scaling exponent κ and the localization length exponent γ indepen- mentary Note 1)28. Figure 1b shows measurements of the longitudinal
dently over several integer-to-integer, integer-to-fractional, and resistance Rxx and the transverse conductance Gxy versus the Landau
fractional-to-fractional Quantum Hall transitions. Contrary to the level filling factor ν; the measurements were performed at B = 13 T,

Fig. 1 | FQH in Bernal-stacked TLG. a Device schematic of TLG encapsulated MLG LLs are marked by the subscripts M, and orbital contents are given by the
between two hBN and few-layer graphite flakes. b Line plots of Gxy (left-axis; solid red numbers 0) band are indicated schematically. d Calculated Landau levels as a func-
line) and Rxx (right-axis; solid blue line) versus ν measured at B = 13 T, T = 20 mK, and tion of energy E and B for D = 0 V/nm. The blue lines are the monolayer-like LLs, while
D = 0 V/nm. The dashed vertical lines mark the FQH states formed at corresponding the red lines are the bilayer-like LLs. The solid and dotted lines indicate the LLs from K
ν, and the arrows indicate corresponding plateaus in Gxy. c Calculated band structure and K 0 -valley, respectively. The solid-green line is the spin-degenerate NM = 0−↑ and
of Bernal stacked trilayer graphene for D = 0 V/nm. The four LLs of the NM = 0 (The NM = 0−↓ monolayer-like LLs that host the FQH states probed in this article.

Nature Communications | (2024)15:8535 2

Article https://doi.org/10.1038/s41467-024-52927-w

T = 20 mK, and D = 0 V/nm. We identify several major odd denomi- To check the validity of our scaling analysis, we perform the fol-
nator FQH states by prominent dips in Rxx and corresponding plateaus lowing error analysis: The residue in the least square fit between the
in Gxy. Indications of developing ν = 3 + 1/5 and 3 + 2/7 states are also scaling curves (like those shown in Fig. 2e) for each assumed value of κ
seen. Several of these FQH states are resolved at B = 4.5 T, attesting to is calculated. This quantity, which we call fit error’, is presented
the high quality of the device in terms of excellent homogeneity of in Supplementary Information Supplementary Fig. 6 and Supplemen-
number density and suppression of long-range scattering (Supple- tary Fig. 7 in a semi-log scale; we find that the fit error is indeed
mentary Information, Supplementary Note 6). minimum for κ = 0.41.
The band structure of TLG is formed of monolayer-like and Figure 3a compiles our findings. These results indicate a κ value
bilayer-like Landau levels (Fig. 1c)—these are protected from mixing by of 0.41 ± 0.03 uniformly observed across all probed transitions
the lattice mirror-symmetry29. The calculated LL spectrum as a func- between IQH and FQH states (compare with Supplementary Fig. 14
tion of B and energy E is shown in Fig. 1d, where blue (red) lines mark of Supplementary Information). This consistency in scaling expo-
the monolayer-like (bilayer-like) LLs. For B > 8 T, the ν = 2 and ν = 3 nents spans various transition types, including (1) transitions from
arise from the spin-split NM = 0−↑ and NM = 0−↓ bands of the one IQH state to another, (2) transitions among different FQH states,
monolayer-like LLs. Here, (+ , −) refers to the two valleys, and (↑, ↓) and (3) transitions between an IQH state and a neighboring FQH
refers to electronic spins. We confine our study to 8 T < B < 13 T to state. It is important to emphasize that the observed universality of κ
avoid Landau level-mixing at lower B and phase transitions between goes beyond marking an experimental confirmation of a uniform
competing FQH states at higher B30–32. scaling law across FQH transitions in any material. Given the distinct
physics of IQH and FQH states, such constancy of the scaling expo-
Critical exponents near FQH plateau-to-plateau nent is remarkable and underscores the universal applicability of the
transitions scaling principle across QH transitions. This is the central result of
Figure 2a shows the T-dependence of Rxy between the IQH states ν = −2 this article.
and ν = −1. Similar data for transition between the FQH states ν = 2 + 2/3
and ν = 2 + 3/5 are shown in Fig. 2b. The critical points νc of the plateau- Locating the transition
to-plateau transition (identified as the crossing point of the Rxy curves The physics of the FQH effect of electrons at a filling factor ν can be
at different T) are indicated in the plots. The exponent κ evaluated mapped onto that of IQH of CF at a filling factor νCF, with ν = νCF/
from the peak value of dRxy/dν versus T near criticality (Fig. 2c and d) in (2νCF ± 1)33. It follows that the critical points for the transition between
both cases is κ = 0.41 ± 0.01. Analysis of the T-dependence of the successive FQH phases at ν = νCF/(2νCF ± 1) and ν = (νCF + 1)/(2(νCF + 1)
inverse of the half-width of Rxx as ν is varied between two consecutive ±1) occur at14,34:
FQH plateaus also yields κ = 0.41 ± 0.02 (Supplementary Information,
Supplementary Note 2). ðν CF + 0:5Þ
νc = : ð3Þ
To demonstrate the scaling properties of Rxy in the vicinity of νc, 2ðν CF + 0:5Þ ± 1
we use the following form13:
The experimentally obtained values of νc, extracted either from the
Rxy ðν,TÞ = Rxy ðν c Þ f ½αðν  ν c Þ ð2Þ crossing point of the Rxy isotherms or the maxima of Rxx, match
exceptionally well with the theoretical predictions (Fig. 3b) (Supple-
0
with α ∝ T−κ. Here, f(0) = 1, and f ð0Þ ≠ 0. This gives us a third, inde- mentary Information, Supplementary Table 1).
pendent method of estimating κ. Figure 2e shows the plots of
Rxy/Rxy(νc) at various temperatures as a function of α∣ν−νc∣ for the Robustness of the critical exponents against LL mixing
ν = 2 +1/3 to 2 + 2/5 transition. α(T) is optimized to collapse the var- A non-zero vertical displacement field D gives rise to a complex phase
ious constant-temperature data onto a single curve (the diagram in TLG, with the Landau levels inter-crossing multiple times,
upper branch of which is for ν < νc and the lower branch is for resulting in significant LL mixing as either D or B is varied35–39. LL-
ν > νc). From the plot of α versus T (inset of Fig. 2e) we obtain mixing can change the effective interaction between the electrons32.
κ = 0.40 ± 0.03. However, as shown in Fig. 3c, it does not significantly affect the

u104
(a) (c) 12 Q = -1 to -2 (e) 1.005 0.0
Q= -1
|dRxy/dQ_Q=Qc (:)

24
9
ln(T)

-0.4
Rxy (k:)

0.24 K -0.8
18 1.60 K
6
3.00 K Qc
0.0 ln(D) 0.3
Q= - 2 N= 0.40 ± 0.01
Rxy/Rxy(Qc

(b) -2.00 -1.50 -1.00 (d) u103 1 10 1.000

10.0 10
Q =8/3 to 13/5 7 to 12
|dRxy/dQ_Q=Qc (:)

Q = 13/5 8
9.9 3 5
Rxy (k:)

0.10 K
6 N = 0.40 ± 0.03
9.8 0.40 K Q
c
0.65 K
1.00 K Q= 8/3
9.7 N= 0.42 ± 0.01 0.4 K 1.3 K
4
0.995
2.58 2.61 2.64 2.67 0.2 0.6 1 2 10-4 10-3 10-2
Q T (K) _QQc|D

Fig. 2 | Scaling near ν = νc. Plot of Rxy versus ν for transition between the (a) IQH states between FQH states 2 + 2/3 and 2 + 3/5. e Scaling analysis of Rxy for the PT transition
ν = −2 and ν = −1 (the critical point νc = −1.5) and b the FQH states 2 + 2/3 and 2 + 3/5 between ν = 2 + 1/3 and ν = 2 + 2/5. The inset is a plot of T versus α in a double loga-
(νc = 2.625). c Double logarithmic plot of ∣dRxy/dν∣ versus T for the PT ν = −2 and ν = −1 rithmic scale (open circles); a linear fit to the data (dotted line) yields κ = 0.40 ± 0.03.
at νc. The dashed line is the fit to the data points using Eq. (1). d Same as in (c) for the PT (For an error analysis, see Supplementary Information, Supplementary Note 7).

Nature Communications | (2024)15:8535 3

Article https://doi.org/10.1038/s41467-024-52927-w

(a) 0.52

/7
24
0.48

3
-3

8/

to
/5
to

to
3

/3

/5
12
-1
-4

7/
0.44

/5

10

17
2

to

13
to
to
N

to

to
3
2
-2

7/
1

3
0.40

/5
0.36

17
to
/3
0.32

10
1.0 1.5 2.0 2.5 3.0 3.5
Qc
(b) (c) 0.52
Rxx maxima
3.5 Rxy minima 0.48

0.44
3.0
Qc

N
0.40

0.36
2.5
0.32
2.5 3.0 3.5 0.0 0.1 0.2 0.3
Qct D (V/nm)

Fig. 3 | Scaling exponents for different PT. a Plot of κ as a function of νc corre- are the values determined from crossing points of isotherms in Rxy, while the circles
max
sponding to different PT evaluated from the maxima of derivative ðdRxy =dνÞ are determined from the Rxx maxima. The black dashed line fits the data points with
near the critical point. The dotted vertical lines mark the experimentally obtained slope = 1.00 ± 0.002. c Plot of κ versus D for the FQH transition from ν = 8/3
νc. The light blue symbols are for the κ values obtained for trilayer graphene, and to ν = 13/5 states evaluated from the maxima of derivative ðdRxy =dνÞmax near
the red symbols are for the single-layer graphene. b Plot of experimentally obtained the critical point. The error bars are determined from the least-square fits to
values of critical points, νc versus those theoretically calculated νct14. The triangles the data.

(a) u10-2 (b) u10-3

3 17 GQ =0.005
10 10 0.2 K
TGxx (K.e2/h)

3 0.2 K 5
2
Gxx (e /h)

1K
1
2

1 GQ =0.015
1.70 K
0 0.1
3.34 3.36 3.38 3.40 3.42 0.8 1.2 1.6 2.0
(c) Q (d) 1/T1/2 (K-1/2)
12 103
J = 2.21r0.07
Gxx/s (e2/h . K)

8 GQ = 0.006 0.015
To (K)

102
4

101

0.008 0.012 0.016 0.005 0.010 0.015

|Q-QC| s1/2 (K-1/2)

Fig. 4 | Scaling exponent in the ES regime for ν = 3 + 2/5 to 3 + 3/7 transition. with an ellipse). c Plots of T0 versus δνCF. The dotted line is a linear fit to the data
a Plots of the T-dependence of Gxx versus filling factor ν for two FQH states between (see Eq. (5)). The slope yields the value of γ. The error bars are determined from the
ν = 3 and ν = 4. The black box marks the region where the ES analysis was carried least square fits to the data in (b). d Plot of scaled longitudinal conductance Gxx/s as
out. b Fit of ES Eq. (4) (dotted lines) to the Gxx data for the transition from ν = 3 + 2/5 a function of scaling parameter s = ∣δν∣γ/T for PT between ν = 17/5 and ν = 24/7. The
and ν = 3 + 3/7. Each set of data points is for a given value of δν=∣ν–νc∣ with νc = 3.416. scatter points of different colors are for different values of ∣δν∣, and the solid black
The plots deviate from the expected ES behavior at high T (the region is marked line is fit to Eq. (6).

universality of κ. This vital result suggests that as long as the anyons dependence of longitudinal conductance Gxx40,41:
are weakly interacting, the critical behavior of the localization-
1=2
delocalization transition remains unaltered. Gxx = G0 eðT 0 =TÞ ð4Þ

Measurement of localization exponent γ with

We now focus on the localized regime, far away from Ec, marked by the
black rectangle in Fig. 4a. Given the presence of strong interactions, it k B T 0 / jδνjγ : ð5Þ
is reasonable to assume that transport in this localized part of the
energy spectrum proceeds through Efros–Shklovskii (ES) type hop- The pre-factor G0 ∝ 1/T and δν = (ν−νc). Figure 4b shows plots of
ping mechanism40. The localization exponent γ determines the T log(TGxx) versus T−1/2 at different values of δν; the dotted lines are linear

Nature Communications | (2024)15:8535 4

Article https://doi.org/10.1038/s41467-024-52927-w

fits to the data. The linearity of the data at low-T is consistent with the value of κ only in devices where long-range scattering dominates.
transport by the ES hopping mechanism in the FQH regime (Eq. (4)). At To our knowledge, ours is the first definite scaling analysis of the QPT
high-T (in the region marked in Fig. 4b by a dotted ellipse), the values over a series of fractional QH states.
of Gxx become relatively large, and the plots deviate from a straight FQH phases are underlined by strongly correlated and interacting
line. In passing, we note that as we move progressively closer to the electrons. Our results demonstrate a surprising correspondence
center of the plateau in Rxy, where the value of Gxx ≈ 0 at low-T, the between the FQH phase transitions and those of non-interacting
linearity of the plots persists to higher temperatures. Fitting T0 electrons. The results indicate a super-universality in the
(estimated from Eqn (4)) and ∣δνCF∣ to Eq. (5), we find the estimated γ to localization–delocalization transitions across distinct anyonic species
lie in the range 2.3−2.6 (Fig. 4c) for FQH plateau-to-plateau transitions, that represent the characteristic quasiparticles of the FQH phases.
very close to the predicted range of γ = 2.3−2.514. The fact that the While much is known about the localization of electrons, the observed
exponent controlling the divergence of the localization length at super universality motivates the study of localization in anyonic qua-
criticality is almost identical for both FQH and IQH states points to an siparticles and the mechanism that drives their conduction in the
effective model of localization that is universal across the different presence of disorder and quasiparticle interactions. Our study raises
statistics of the quasiparticles in these QH phases. Furthermore, from the natural question of whether the universality observed in this con-
κ = 1/zγ ≈ 0.41 ± 0.03 and γ ≈ 2.3, we get z ≈ 1, as expected for a strongly text applies to transitions between other topological phases with
interacting system7,9,42,43. fractional excitations, such as fractional Chern insulators45.
An independent estimate of γ is obtained by casting Eq. (4) into a
single-parameter scaling form44: Methods
Device fabrication
1=2
Gxx ðsÞ = σ*seðT*sÞ , ð6Þ Devices of dual graphite gated ABA trilayer graphene (TLG) hetero-
structures were fabricated using a dry transfer technique (for details,
with the the scaling parameter s = ∣δνCF∣γ/T. Figure 4d shows the scaling see Supplementary Information Supplementary Note 1). Raman spec-
plots of Gxx/s versus s1/2 for the PT in ES regime from ν = 3 + 2/5 to troscopy and optical contrast were used to determine the number of
ν = 3 + 3/7. We find a near-perfect data collapse for all values of δνCF in layers and the stacking sequence. The devices were patterned using
the localized regime with γ ≈ 2.3, providing an independent validation electron beam lithography followed by reactive ion etching and thermal
of the universality of γ. deposition of Cr/Pd/Au contacts. Dual electrostatic gates were used to
simultaneously tune the areal number density n = [(CtgVtg + CbgVbg)/
Discussion e + no] and the displacement field D = [(CbgVbg−CtgVtg)/2ϵ0 + D0] across
We are now in a position to compare the universality of κ seen in the the device. Here Cbg(Ctg) is the capacitance of the back gate (top gate),
FQH PT in our high-mobility TLG with non-universality of the same and Vbg(Vtg) is the voltage of the back gate (top gate). The values of Ctg
measured in the high-mobility 2D semiconductors23. The large and Cbg are determined from quantum Hall measurements. no and Do
spread in the observed values of κ seen in the data in GaAs quantum are the residual number density and electric field due to unavoidable
wells was attributed to two main reasons23. The first is the formation impurities in the channel.
of numerous emerging FQH phases between ν = 1/3 and 2/5, which
limits the temperature range over which one observes the decrease Transport measurements
of the width of Rxx with T. Note that in Fig. 1b, there are two incipient The electrical transport measurements were performed in a dilution
FQH phases, ν = 3 + 1/5 and 3 + 2/7, between the more robust phases refrigerator (with a base temperature of 20 mK) at low frequency
ν = 3 and ν = 3 + 1/3. The incipient phases are weak enough not (11.4 Hz) using standard low-frequency measurement techniques, with
to affect the scaling of the transition region in Rxy even at the lowest a bias current of 10 nA.
temperature employed here. As a result, we find κ = 0.42 ±
0.01 (Fig. 3a). Data availability
The second reason is related to the type of disorder in the The authors declare that the data supporting the findings of this study
sample23. Universality in κ is observed only when the effective disorder are available within the main text and its Supplementary Information
potential is short-ranged20, as in our graphite-gated high-mobility and at https://doi.org/10.6084/m9.figshare.26809147.v1. Other rele-
graphene devices. This is not the case in GaAs/AlGaAs systems, where vant data are available from the corresponding author upon request.
long-range scattering potential from the impurities cannot be
ignored23. We fabricated graphene devices without the graphite gate Code availability
electrodes to probe the effect of long-range interactions on κ. The The codes that support the findings of this study are available from the
graphene channel was no longer screened from long-range Coulomb corresponding author upon request.
fluctuations arising from the SiO2 substrate; this was reflected in
reduced mobility ~2−5 m2/Vs. While in these devices, we do not find References
FQH states, the value of κ for IQH transitions varied widely between 1. Klitzing, K. V., Dorda, G. & Pepper, M. New method for high-
0.45−0.64 (Supplementary Information, Supplementary Note 4), sup- accuracy determination of the fine-structure constant based on
porting the conclusions of ref. 23. quantized hall resistance. Phys. Rev. Lett. 45, 494–497 (1980).
To summarize, our principal finding is that scaling properties for 2. Huo, Y. & Bhatt, R. N. Current carrying states in the lowest landau
transitions involving Abelian FQH states and/or IQH phases are uni- level. Phys. Rev. Lett. 68, 1375–1378 (1992).
versal. Specifically, we have demonstrated the scaling of the long- 3. Laughlin, R. B. Quantized hall conductivity in two dimensions. Phys.
itudinal conductance (with a scaling exponent κ = 0.41 ± 0.02 and Rev. B 23, 5632–5633 (1981).
localization exponent γ ≈ 2.3) in the IQH and FQH states in Bernal- 4. Aoki, H. & Ando, T. Critical localization in two-dimensional landau
stacked ABA trilayer graphene. This conclusion holds for plateau-to- quantization. Phys. Rev. Lett. 54, 831–834 (1985).
plateau transitions between two consecutive IQH states, two FQH 5. Chalker, J. T. & Daniell, G. J. Scaling, diffusion, and the integer
states, and even between IQH and the adjoining FQH state, underlining quantized hall effect. Phys. Rev. Lett. 61, 593–596 (1988).
the universal character of the scaling. This universality of κ persists 6. Chalker, J. T. & Coddington, P. D. Percolation, quantum tunnelling
even when an external displacement field hybridizes the Landau levels and the integer hall effect. J. Phys. C: Solid State Phys. 21, 2665
of Bernal-stacked TLG. In fact, we find deviations from universality in (1988).

Nature Communications | (2024)15:8535 5

Article https://doi.org/10.1038/s41467-024-52927-w

7. Huckestein, B. & Backhaus, M. Integer quantum hall effect of 31. Zhu, Z., Sheng, D. N. & Sodemann, I. Widely tunable quantum phase
interacting electrons: Dynamical scaling and critical conductivity. transition from moore-read to composite Fermi liquid in bilayer
Phys. Rev. Lett. 82, 5100–5103 (1999). graphene. Phys. Rev. Lett. 124, 097604 (2020).
8. Halperin, B. I. & Hohenberg, P. C. Scaling laws for dynamic critical 32. Sodemann, I. & MacDonald, A. H. Landau level mixing and the
phenomena. Phys. Rev. 177, 952–971 (1969). fractional quantum Hall effect. Phys. Rev. B 87, 245425 (2013).
9. Li, W. et al. Scaling in plateau-to-plateau transition: a direct con- 33. Jain, J. K. Composite-fermion approach for the fractional quantum
nection of quantum Hall systems with the Anderson localization hall effect. Phys. Rev. Lett. 63, 199–202 (1989).
model. Phys. Rev. Lett. 102, 216801 (2009). 34. Goldman, V. J., Jain, J. K. & Shayegan, M. Nature of the extended
10. Sondhi, S. L., Girvin, S. M., Carini, J. P. & Shahar, D. Continuous states in the fractional quantum hall effect. Phys. Rev. Lett. 65,
quantum phase transitions. Rev. Mod. Phys. 69, 315–333 (1997). 907–910 (1990).
11. Pruisken, A. M. M. Universal singularities in the integral quantum 35. Zibrov, A. A. et al. Emergent Dirac gullies and gully-symmetry-
hall effect. Phys. Rev. Lett. 61, 1297–1300 (1988). breaking quantum hall states in a b a trilayer graphene. Phys. Rev.
12. Dodoo-Amoo, N. A. et al. Non-universality of scaling exponents in Lett. 121, 167601 (2018).
quantum hall transitions. J. Phys.: Condens. Matter 26, 36. Rao, P. & Serbyn, M. Gully quantum Hall ferromagnetism in biased
475801 (2014). trilayer graphene. Phys. Rev. B 101, 245411 (2020).
13. Huckestein, B. Scaling theory of the integer quantum hall effect. 37. Winterer, F. et al. Spontaneous gully-polarized quantum hall states
Rev. Mod. Phys. 67, 357–396 (1995). in aba trilayer graphene. Nano Lett. 22, 3317–3322 (2022).
14. Pu, S., Sreejith, G. J. & Jain, J. K. Anderson localization in the frac- 38. Serbyn, M. & Abanin, D. A. New dirac points and multiple landau
tional quantum hall effect. Phys. Rev. Lett. 128, 116801 (2022). level crossings in biased trilayer graphene. Phys. Rev. B 87,
15. Huckestein, B. & Kramer, B. One-parameter scaling in the lowest 115422 (2013).
landau band: Precise determination of the critical behavior of the 39. Wang, Yun-Peng, Li, Xiang-Guo, Fry, J. N. & Cheng, Hai-Ping First-
localization length. Phys. Rev. Lett. 64, 1437–1440 (1990). principles studies of electric field effects on the electronic structure
16. Tsui, D. C., Stormer, H. L. & Gossard, A. C. Two-dimensional mag- of trilayer graphene. Phys. Rev. B 94, 165428 (2016).
netotransport in the extreme quantum limit. Phys. Rev. Lett. 48, 40. Efros, A. L. & Shklovskii, B. I. Coulomb gap and low temperature
1559–1562 (1982). conductivity of disordered systems. J. Phys. C: Solid State Phys. 8,
17. Hohenberg, P. C. & Halperin, B. I. Theory of dynamic critical phe- L49 (1975).
nomena. Rev. Mod. Phys. 49, 435–479 (1977). 41. Ono, Y. Localization of electrons under strong magnetic fields in a
18. Thouless, D. J., Kohmoto, M., Nightingale, M. P. & den Nijs, M. two-dimensional system. J. Phys. Soc. Jpn. 51, 237–243 (1982).
Quantized hall conductance in a two-dimensional periodic poten- 42. Hohls, F. et al. Dynamical scaling of the quantum Hall plateau
tial. Phys. Rev. Lett. 49, 405–408 (1982). transition. Phys. Rev. Lett. 89, 276801 (2002).
19. Wei, H. P., Lin, S. Y., Tsui, D. C. & Pruisken, A. M. M. Effect of long- 43. Polyakov, D. G. & Shklovskii, B. I. Conductivity-peak broadening in
range potential fluctuations on scaling in the integer quantum hall the quantum Hall regime. Phys. Rev. B 48, 11167–11175 (1993).
effect. Phys. Rev. B 45, 3926–3928 (1992). 44. Hohls, F., Zeitler, U. & Haug, R. J. Hopping conductivity in the
20. Li, W., Csáthy, G. A., Tsui, D. C., Pfeiffer, L. N. & West, K. W. Scaling quantum Hall effect: revival of universal scaling. Phys. Rev. Lett. 88,
and universality of integer quantum hall plateau-to-plateau transi- 036802 (2002).
tions. Phys. Rev. Lett. 94, 206807 (2005). 45. Han, T. et al. Correlated insulator and Chern insulators in pentalayer
21. Engel, L., Wei, H. P., Tsui, D. C. & Shayegan, M. Critical exponent in rhombohedral-stacked graphene. Nat. Nanotechnol. 19, 181–187
the fractional quantum hall effect. Surf. Sci. 229, 13–15 (1990). (2024).
22. Machida, T., Ishizuka, S., Komiyama, S., Muraki, K. & Hirayama, Y.
Scaling in fractional quantum hall transitions. Phys. B: Condens. Acknowledgements
Matter 298, 182–186 (2001). We thank Jainendra K. Jain, Sankar Das Sarma, Rajdeep Sensarma,
23. Madathil, P. T. et al. Delocalization and universality of the fractional Nandini Trivedi, Ravindra Bhatt and Prasant Kumar for helpful discus-
quantum hall plateau-to-plateau transitions. Phys. Rev. Lett. 130, sions and clarifications. We acknowledge Ramya Nagarajan for data on
226503 (2023). low-mobility samples. A.B. acknowledges funding from U.S. Army
24. Kumar, P., Nosov, P. A. & Raghu, S. Interaction effects on quantum DEVCOM Indo-Pacific (Project number: FA5209 22P0166) and Depart-
hall transitions: dynamical scaling laws and superuniversality. Phys. ment of Science and Technology, Govt of India (DST/SJF/PSA-01/2016-
Rev. Res. 4, 033146 (2022). 17). K.W. and T.T. acknowledge support from the JSPS KAKENHI (Grant
25. Pan, W. et al. Particle-hole symmetry and the fractional quantum Numbers 21H05233 and 23H02052) and World Premier International
hall effect in the lowest landau level. Phys. Rev. Lett. 124, Research Center Initiative (WPI), MEXT, Japan. G.J.S. thanks Condensed
156801 (2020). Matter Theory Center and Joint Quantum Institute, University of Mary-
26. Sarkar, S. et al. Role of different scattering mechanisms on the land College Park, for their hospitality during the preparation of this
temperature dependence of transport in graphene. Sci. Rep. 5, manuscript. Y.G. acknowledges the support by the Deutsche For-
16772– (2015). schungsgemeinschaft (DFG) through grant No. MI 658/10-2 and RO
27. Rhodes, D., Chae, SangHoon, Ribeiro-Palau, R. & Hone, J. Disorder in 2247/11-1, the US-Israel Binational Science Foundation 2022391, and the
van der Waals heterostructures of 2d materials. Nat. Mater. 18, Minerva Foundation. Y.G. is the incumbent of the InfoSys chair at IISc.
541–549 (2019).
28. Pizzocchero, F. et al. The hot pick-up technique for batch Author contributions
assembly of van der Waals heterostructures. Nat. Commun. S.K., T.C., K.R.A., D.S., and A.B. conceived the idea of the study, con-
7, 11894 (2016). ducted the measurements, and analyzed the results. T.T. and K.W. pro-
29. Koshino, M. & McCann, E. Landau level spectra and the quantum vided the hBN crystals. U.G., G.J.S., and Y.G. developed the theoretical
Hall effect of multilayer graphene. Phys. Rev. B 83, 165443 (2011). model. All the authors contributed to preparing the manuscript.
30. Papić, Z., Abanin, D. A., Barlas, Y. & Bhatt, R. N. Tunable interactions
and phase transitions in Dirac materials in a magnetic field. Phys. Competing interests
Rev. B 84, 241306 (2011). The authors declare no competing interests.

Nature Communications | (2024)15:8535 6

Article https://doi.org/10.1038/s41467-024-52927-w

Additional information Open Access This article is licensed under a Creative Commons
Supplementary information The online version contains Attribution-NonCommercial-NoDerivatives 4.0 International License,
supplementary material available at which permits any non-commercial use, sharing, distribution and
https://doi.org/10.1038/s41467-024-52927-w. reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the
Correspondence and requests for materials should be addressed to Creative Commons licence, and indicate if you modified the licensed
Aveek Bid. material. You do not have permission under this licence to share adapted
material derived from this article or parts of it. The images or other third
Peer review information Nature Communications thanks the anon- party material in this article are included in the article’s Creative
ymous, reviewer(s) for their contribution to the peer review of this work. Commons licence, unless indicated otherwise in a credit line to the
A peer review file is available. material. If material is not included in the article’s Creative Commons
licence and your intended use is not permitted by statutory regulation or
Reprints and permissions information is available at exceeds the permitted use, you will need to obtain permission directly
http://www.nature.com/reprints from the copyright holder. To view a copy of this licence, visit http://
creativecommons.org/licenses/by-nc-nd/4.0/.
Publisher’s note Springer Nature remains neutral with regard to jur-
isdictional claims in published maps and institutional affiliations. © The Author(s) 2024

Nature Communications | (2024)15:8535 7