VOLUME 50, +UMBER I8 PHYSICAL REVIEW LETTERS 2 MAv 1983

Anomalous Quantum Hall Effect: An Incompressible Quantum Fluid

with Fractionally Charged Excitations
H. B. Laughlin
Iasorence Livermore National Laboratory, University of California, Jivermore, California 94550
(Received 22 February 1983)
This Letter presents variational ground-state and excited-state wave functions which
describe the condensation of a two-dimensional electron gas into a new state of matter.
PACS numbers: 71.45.Nt, 72.20.My, 73.40. Lq

The "—, '" effect,

recently discovered by Tsui, sistent with all the experimental facts and ex-
Stormer, and Gossard, ' results from the conden- plain the effect. The ground state is a new state
sation of the two-dimensional electron gas in a of matter, a quantum fluid the elementary exci-
GaAs-Ga„Al, „As heterostructure into a new type tations of which, the quasielectrons and quasi-
of collective ground state. Important experimen- holes, are fractionally charged. I have verified
tal facts are the following: (1) The electrons con- the correctness of these wave functions for the
dense at a particular density, —, ' of a full Landau case of small numbers of electrons, where direct
level. (2) They are capable of carrying electric numerical diagonalization of the many-body Ham-
current with little or no resistive loss and have a iltonian is possible. I predict the existence of a
Hall conductance of —,'e'/h. (3) Small deviations sequence of these ground states, decreasing in
of the electron density do not affect either con- density and terminating in a Wigner crystal.
ductivity, but large ones do. (4) Condensation Let us consider a two-dimensional electron
occurs at a temperature of —1.0 K in a magnetic gas in the x-y plane subjected to a magnetic field
field of 150 kG. (5) The effect occurs in some H, in the z direction. I adopt a symmetric gauge
samples but not in others. The purpose of this vector potential A = ,'H, [xy-yx—] and write the
Letter is to report variational ground-state and eigenstates of the ideal single-body Hamiltonian
excited-state wave functions that I feel are con- —j (h/i)V —(e/c)A ' in the manner
H, p — l

I
I

jm, n) =(2 '""~m!n!) 'i'exp[-;(x'+y')] —+i — ——i — exp [ —ggx'+y')]

Bg Bg
m
B
Bx B$
5

with the cyclotron energy he, = h(equi, /mc) and

the magnetic length a, =(A/m~, )' '=(Ac/eH, )' ' states of the lowest Landau level as
set to 1. We have
') Im) =(2 +~em!) "i g exp( —« jg j )
e, p l m, n) = (n+ —,
l m, n). (2)
The manifold of states with energy n+ & consti- where z=x+ iy. m) is an eigenstate of angular
l

tutes the nth Landau level. I abbreviate the momentum with eigenvalue m. The many-body
Hamiltonian is

jI = g { (fi /i )V, —(e /c)A, ' + V(z, )j + Z e'/ ~, - z, I,

j j j

where jand 4 run over the N particles and V is a

j
m about the center of mass are of the form
potential generated by a uniform neutralizing
background. -~.)
&=(~ (~+~,)"exp[- l(j~ j'+ jz. j')] (»
I showed in a previous paper' that the —,' effect
could be understood in terms of the states in the My present theory generalizes this observation to
lowest I andau level solely. With e'/a, s h ~„ A' particles.
the situation in the experiment, quantization of I write the ground state as a product of Jastrow
interelectronic spacing follows from quantization functions in the manner
of angular momentum: The only wave functions 0= II f(z, —z, )jexp(- 'Q,
composed of states in the lowest Landau level
lz, l'),
—, (6)
which describe orbiting with angular momentum and minimize the energy with respect to f. We

1983 The American Physical Society 1395

VOLUME 50, +UMBER 18 PHYSICAL REVIEW LETTERS 2 MAY 1983

observe that the condition that the electrons lie TABLE I. Projection of variational three-body wave
in the lowest Landau level is tha. t f(z) be poly- functions (|) in the manner (g C )/((( ) (4
~ ~
(l)

nomial in z. The antisymmetry of g requires that x c ~)) (/'. 4 is the lowest-energy eigenstate of angu-
lar momentum 3m calculated with V = 0 and an inter-
f be odd. Conservation of angular momentum re-
electronic potential of either 1/x, -In(x), or exp(-x'/
quires that II, ,„f(z, —z~) be a homogeneous poly- 2).
nomial of degree M, where M is the total angular
momentum. We have, therefore, f(z) =z, with exp(- x'/2)
m odd. To determine which m minimizes the en-
ergy, I write 1 1 1 1
3 0. 999 46 0.996 73 0.999 66
14.1' = I{II,„(z,-z, ) kexp(- -'Z, lz, l')I' 5 0. 994 68 0.991 95 0.999 39
(7) 7 0.994 76 0.992 95 0.99981
9 0.995 73 0.994 37 0.999 99
where P= 1/m and 4 is a classical potential ener- ll 0.996 52 0.995 42 0.999 96
gy given by
13 0.997 08 0.996 15 0.999 85
4 =-Z, ,„2m'»lz, -z, I+ am', lz, I' (8)
within a few percent by the ion disk energy:
4 describes a system of N identical particles of
charge Q = m, interacting via logarithmic poten- U Ot gm —d2y+~
2

j+j
0 2
2
——
j+
e2
dy 2' 2

tials and embedded in a uniform neutralizing j

(10)
background of charge density o =(27)a,') '. This = (4/37) —1)2e'/R,
is the classical one-component plasma (OCP), a where the integration domain is a disk of radius
system which has been studied in great detail. 8 = (pro ) "' '. At I' = 2 we have the exact result'
Monte Carlo calculations' have indicated that the that g(r) =1 —exp[- (r/fI)'], giving U„, = ——, ')T'/'e'/
OCP is a hexagonal crystal when the dimension- At m=3 and m=5 I have reproduced the Monte
less plasma parameter I'= 2''= 2m is greater Carlo g(r) of Caillol et al. ' using the modified
than 140 and a fluid otherwise. ~ (
~' describes a hypernetted chain technique described by them.
system uniformly expanded to a density of o I obtain U„, =(-0.4156+0.0012)e'/a, and U„,(5)=
=m '(27)a, ') '. lt minimizes the energy when o .
(-0.3340 +0.0028)e'/a, The corresponding values
equals the charge density generating V. for the charge-density wave4 are —0. 389e'/ao and
In Table I, I list the projection of g for three —0. 322e'/ao. U„, is a smooth function of I I .
particles onto the lowest-energy eigenstate of interpolate it crudely in the manner
angular momentum 3m calculated numerically.
These are all nearly 1. This supports my asser- U tot (m) =-'0.— ——
814 —
0. 230
064 -1
—
v 0
tion that a wave function of the form of Eq. (6) has
This interpol. ation converges to the CD% energy
adequate variational freedom. I have done a sim-
near rn = 10. The actual crystallization point can-
ilar calculation for four particles with Coulombic
not be determined from that of the OCP since the
repulsions and find projections of 0.979 and 0.947
CD% has a lower energy than the crystal de-
for the m = 3 and m = 5 states.
has a total energy per particle which for
scribed by (
for m& 71 ~

small m is more negative than that of a charge-

I generate the elementary excitations of g by
density wave (CDW). ' It is given in terms of the
piercing the fluid at z, with an infinitely thin
solenoid and passing through it a flux quantum
radial distribution function g(r) of the OCP by
t) cp =- hc/e adiabatically. The effect of this opera-
oo tion on the single-body wave functions is
U„,= ~ —[g(r) —I ] rdr. (9)
(z-z, ) exp(-4lzl')-(z-z, ) "exp(--'. ~z~'). (»)
In the limit of large I, U„, is approximated Let us take as approximate representations of
these excited states
(13)

'4 "=&., ''4= "(--'Xl, (, l')In;(, ——I(n;, .(*, —,)"),

~o
". (14)

1396
VOLUME 50, +UMBER 18 PHYSICAL REVIEW LETTERS 2 MAv 1983

for the quasihole and quasieleetron, respectively. For m--3, these estimates are 0.062e'/a, and
For four particles, I have projected these wave 0.038e'/a, . This compares well with the value
functions onto the analogous ones computed nu- 0.033e'/ao estimated from the numerical four-
merically. I obtain 0. 998 for g, ' and 0. 994 for
" "
I obtain 0. 982 for g, =fg, (z, -Z))g„
partiele solution in the manner
")
~ =-.'(E(q. -') E(e. —2E(y, )), (18)
which is P, with the center-of-mass motion re-
moved. where E(g, ) denotes the eigenvalue of the numer-
These excitations are particles of charge 1/m. ical analog of P, . This expression averages the
To see this let us write [ (+'oI' as e s ~, with electron and hole creation energies while sub-
P=1/m and tracting off the error due to the absence of V.
I have performed two-component hypernetted
O' = 4 —2+, In~ z, —zo(. (15) chain calculations for the energies of P, "o and
4' describes an QCP interacting with a phantom I obtain (0.022 a 0.002)e'/ao and (0.025
point charge at z, . The plasma will completely + 0.005)e'/a, . If we assume a value e = 13 for
screen this phantom by accumulating an equal and the dielectric constant of GaAs, we obtain 0.02e'/
opposite charge near zp. However, since the cap ~4 K when Ho = 150 kG.
plasma in reality consists of particles of charge The energy to make a particle does not depend
1 rather than charge rn, the real accumulated on z„so long as its distance from the boundary
charge is 1/m. Similar reasoning applies to '0
t/r is greater than its size. Thus, as in the single-
if we approximate it as g, (z, —z, ) 'P, g„where partiele problem, the states are degenerate and
P, is a projection operator removing all con- there is no kinetic energy. We can expand the
figurations in which any electron is in the single- creation operator as a power series in z, :
body state (z- z, )'exp(- 4 Iz I'). The projection of
z, )z, '- .
E
this approximate wave function onto g, 'o for four gX, (,
~, = j=0 ~
~,
particles is 0. 922. More generally, one observes
that far away from the solenoid, adiabatic addi- These A. , are the elementary symmetric poly-
tion of Ay moves the fluid rigidly by exactly one nomials, ' the algebra of which is known to span
state, per Eq. (12). The charge of the particles the set of symmetric functions. Since every anti-
is thus 1/m by the Schrieffer counting argument. ' symmetric function ean be written as a sym. -
The size of these particles is the distance over metric function times y„ these operators and
which the QCP screens. Were the plasma weakly their adjoints generate the entire state space.
coupled (I' & 2) this would be the Debye length AD= It is thus appropriate to consider them X lin-
a, /N. For the strongly coupled plasma, a better early independent particle creation operators.
estimate is the ion-disk radius associated with a The state described by g is incompressible
charge of 1/m: A=&2a, . From the size we can because compressing or expanding it is tanta-
estimate the energy required to make a particle. mount to injecting particles. If the area of the
The charge accumulated around the phantom in system is reduced or increased by ~A the en-
the Debye-Huckel approximation is ergy rises by &U = o ss M ~. Were this an elast-
~

ic solid characterized by a bulk modulus B, we

would have 5U= —, 'B(5A)'/A. Incompressibility
causes the longitudinal collective excitation
where Ko is a modified Bessel function of the roughly equivalent to a compressional sound
second kind. The energy required to accumulate wave to be absent, or more precisely, to have
it is an energy - 4 in the long-wavelength limit. This
e' facilitates current conduction with no resistive
bp5p v 1
(16) loss at zero temperature. Our prototype for
this behavior is full Landau level (m = 1) for
This estimate is an upper bound, since the plasma which this collective excitation occurs at R~, .
is strongly coupled. To make a better estimate The response of this system to compressive
let 5p = a inside the ion disk and zero outside, to stresses is analogous to the response of a type-
obtain II superconductor to the application of a magnetic
field. The system first generates Hall currents
3 1e' (17)
without compressing, and then at a critical
disk
2v2p m' a p stress collapses by an area quantum ppl2~QO'
1397
VOLUME 50, +UMBER 18 PHYSICAL REVIEW LETTERS 2 MA+ 198$

and nucleates a particle. This, like a flux line, for helpful discussions. I also wish to thank
is surrounded by a vortex of Hall current rotat- P. A. Lee, D. Yoshioka, and B. I. Halperin for
ing in a sense opposite to that induced by the helpful criticism. This work was performed
stress. under the auspices of the U. S. Department of
The role of sample impurities and inhomogenei- Energy by Lawrence Livermore National Labora-
ties in this theory is the same as that in my tory under Contract No. W-7405-Eng-48.
theory of the ordinary quantum Hall effect. ' The
electron and hole bands, separated in the im-
purity-free case by a gap 2A, are broadened
into a continuum consisting of two bands of ex- 'D. C. Tsui, H. L. Stormer, and A. C. Qossard, Phys.
tended states separated by a band of localized Rev. Lett. 48, 1559 (1982).
ones. Small variations of the electron density 'R. B. Laughlin, Phys. Rev. B 27, 3383 (1983).
3J. M. Caillol, D. Levesque, J. J. leis, and J. P.
move the Fermi level within this localized state
Hansen, J. Stat. Phys. 28, 325 (1982).
band as the extra quasiparticles become trapped D. Yoshioka and H. Fukuyama, J. Phys. Soc. Jpn.
at impurity sites. The Hall conductance is (1/m) 47, 394 (1979}; D. Yoshioka and P. A. Lee, Phys. Rev.
& (e'/h) because it is related by gauge invariance B 27, 4986 {1983), and private communication.
to the charge of the quasiparticles e* by o H, z ~B. Jancovici, Phys. Rev. Lett. 46, 386 {1981). -I=-2
=e*e/h, whenever the Fermi level lies in a corresponds to a full Landau level, for which the total
localized state band. As in the ordinary quantum energy equals the Hartree-Fock energy P~/8 e2/ao. -
This correspondence may be viewed as the underlying
Hall effect, disorder sufficient to localize all the
reason an exact solution at 1 = 2 exists.
states destroys the effect. This occurs when the 6W. P. Su and J. R. Schrieffer, Phys. Rev. Lett. 46,
collision time w in the sample in the absence of 738 (1981).
a magnetic field becomes smaller than T &h/4. S. Lang, Algebra (Addison-Wesley, Reading, Mass. ,
I wish to thank H. DeWitt for calling my atten- 1965}, p. 132.
tion to the Monte Carlo work and D. Boercker 'R. B. Laughlin, Phys. Rev. B 23, 5632 (1981).

1398