• Aucun résultat trouvé

Schrödinger operators with form-bounded potentials in <span class="mathjax-formula">$L^p$</span>-spaces

N/A
N/A
Protected

Academic year: 2022

Partager "Schrödinger operators with form-bounded potentials in <span class="mathjax-formula">$L^p$</span>-spaces"

Copied!
12
0
0

Texte intégral

(1)

A NNALES DE L ’I. H. P., SECTION A

M. A. P ERELMUTER

Schrödinger operators with form-bounded potentials in L

p

-spaces

Annales de l’I. H. P., section A, tome 52, n

o

2 (1990), p. 151-161

<http://www.numdam.org/item?id=AIHPA_1990__52_2_151_0>

© Gauthier-Villars, 1990, tous droits réservés.

L’accès aux archives de la revue « Annales de l’I. H. P., section A » implique l’accord avec les conditions générales d’utilisation (http://www.numdam.

org/conditions). Toute utilisation commerciale ou impression systématique est constitutive d’une infraction pénale. Toute copie ou impression de ce fichier doit contenir la présente mention de copyright.

Article numérisé dans le cadre du programme Numérisation de documents anciens mathématiques

http://www.numdam.org/

(2)

151

Schrödinger operators with form-bounded potentials

in Lp-spaces

M. A. PERELMUTER

Urickij street, 11, apt. tS3, Kiev, 252035,

Ann. Inst. Henri Poincaré,

Vol. 52, 2, t990, Physique théorique

ABSTRACT. - Let

closure of

the operator (-

A + V +

C) f C~0 (Rl)

is

generator

of the contrac- tion

Co-semigroup

in For

p = ~

we obtain a new theorem concern-

ing

the essential

self-adjointness

of the

Schrödinger operator.

The

sharp

ness of the results is illustrated

by

the

examples.

ture de est un

generateur

de semi-

groupe

Co

de contraction dans Pour

p = 2

nous obtenons un

Annales de l’Institut Henri Poincaré - Physique theorique - 0246-0211 1 Vol. 52/90/02/151/11/$3.10/ (è) Gauthier-Villars

(3)

152 M. A. PERELMUTER

nouveau théorème concernant

l’auto-adjonction

essentielle de

l’opérateur

de

Schrodinger.

Ces resultats sont illustres sur des

exemples.

~

INTRODUCTION

This article deals with the

Schrodinger operators -A+V acting

in

Lp - LP

(IR’, cf x).

We want to

investigate

the first

spectral problem,

i. e.

the

problem

of

m-accretivity.

Most of the results

concerning

this

problem

contain the condition of - 0394-boundedness of the

operator

V _

==max{-V,0}.

Our

goal

is to

investigate

the conditions of the

following type:

where F is some function space.

The first result in this direction is due to Povzner

[1] ] [p=2, P=l, F=C(!R~)]. Recently

some results were also obtained. In

particular

the essential

self-adjointness

was

proved provided

that

F = Lk,

My

purpose here is to extend the abovementioned results to the case of

LP-spaces

and to choose the broader space F. Even for

p = 2

we obtain a

new result which is very close to

optimal.

It should be noted that our method is an extension of the Semenov’s

one

[3].

Note that LP-A-boundedness

implies

the condition

(1) (see [4])

and in

my

opinion

the condition

( 1 ) corresponds

to the

point

of our

problem.

We want also to

emphasize

that the exact

eigenfunction

estimates for the

operator - A

+ V and the exact results

concerning

the

accretivity

of the

Schrodinger operators

have been

proved

under the same condition

( 1 ) (see [5], [6]).

Annales de l’lnstitut Henri Poincaré - Physique théorique

(4)

SCHRODINGER OPERATORS 153

Some common

symbols

are:

==

x) = the complex functions f’

on

fR’

such that

f

p E

LP,

V (p e meas

(B)

=

Lebesgue

measure of B c

R’ ;

Lp° ~° _--_ LP’ 00

d x)

= the

complex functions f

on

~~

such that

(note

that LP c LP’ 00

and

E

LP’ 00) Lp,~comp

= the set of functions in Lp,~ with

compact support. A= Y 2014_

the

Laplacian acting

on D’

(1R1).

j= 1

lx )

MAIN RESULTS

LEMMA 1. -

If V

E

then ( ~ V ~p - £

= O 0.

Proof. - &#x3E;t}.

The

function (t)

is nonin-

creasing and (0) = meas {supp V}

CIJ. The definition of the space

implies

that

~. ( t) _ C t - p, V t &#x3E; 0,

so

DEFINITION. - Let 0 ~

V,

W E

L1loc.

Then V ~

PK03B2(-0394

+

W)

if and

only

if

Vol. 52, n° 2-1990.

(5)

154 M A. PERELMU"fER

THEOREM 2. - tet V be a real valued measurable

function such

that

Then the range

of

tk

operator (-0394+V+03BB) C~0

M dense M!

LP for

any

X&#x3E;C(P).

Suppose

that the range is not

dense,

then

by

the Hahn-Banach

theorem~

there exists a

function 0 ~ u ~ Lp

such that

Whithout loss we can assume u= Reu. r~ _

~’ j~’.

Let

~&#x3E;0 be a small

parameter

and Iet us

and It follows that

n

L’.

The

equality (3) gives

in the distributional sense; hence

(C£ 17l).

Moreover, implies

Let us consider the

operator 1~Z

with the domain

where

(-A+V+)

is taken in the distributional sense and

==

(-A+V+)~

H:

is the

generator

of the

strongly

continuous contraction

semigroup

== V~&#x3E;0 in LZ which has

properties

such as follows

below

l’Institut Hener Poincaré - Physique théorique

(6)

SCHRÖDINGER OPERATORS 155

Let

03C6r=Ts(l r)u.

h follows from

(5), (8), (4)

and

C0-property

that

According

to

(I)

we can choose

subsequence (which

we shall denote

by

the same

symbol { ~ } )

such that

Then

Indeed,

because

~(jc)-~(j~)

a. c. Hence in order to conclude the

proof

of

(3)

it

is sufficient to prove the convergence of the norms

(a

nice

proof

of this

fact may be found in

(8])

but

and the convergence

U

cp, -

it

u is the consequence of

(~).

ln the same way we obtain

Let

~~=7~ ~ ~ By

weR known

properties

of the

operation ~ ~ (see,

e. g.~

[9D~

we have

(1), (3), (4)

for the

subsequence

of

{ fP,. } ~ (7) implies i,2,...)

so that

V+ p~~V+

qJ,.. This fact and the

equality

L"

A

(/~ ~ p~)=~ ~

show that

Let us choose any

03C9~C~0 having

and let and

It is easy to see that

properties (!), (3), (4)

hoid for

Vol. 52, n* 2-1990.

(7)

156 M. A. PERELMUTER

Using

the

standard diagonal

process we can choose the

subsequence

th. = s uch that t

Observe that so that

M,JM,)~’~ M,~’~e~(V)(~~e.

g.,

[ 1 0],

Chapter 2),

and the direct calculation

yields

the

identity

(4)-(6) imply

the finiteness of the

quantities

so that the

simple approximation arguments

show that

(3)

is valid

provided

that and

(9)-( 1 2 ) imply

Using (13), (14)

and the condition V _ E + V

+ )

we have

:ic° l’Institut Henri Physique theorique

(8)

157 SCHRODINGER OPERATORS

Note

Therefore,

According

to

(!5), (!6).

~ ‘1 1 ...

From (14)

we have that the

right

side of the last

inequality equals

So, using

Holder

inequality

we obtain

According

to the lemma I

IIV - !1z

h:)

= f~’

(E - ~k~.

On the other hand

Now:

assuming

that

E!

0 in

( 1 7)

and.

using k ~ ~

I we obtain

Then u - o as 03BB &#x3E; C

(03B2). []

~~marks. - l . The condition V _ E

F’I~~~( - ~

+ V

+~ ~1

is necessary iw the theorem 2 but it i.s not sufficient as it may be seen in the

example

- 0394 -

03BA 1 - 2 (see [ 11], C’hapter X).

The same

example

shows the

sharpness

~x~2

of the theorem 2.

Inded,

let 2 p m

l,

where

)((.)

is the indicator of the unit ban in Then the range of the operator

(-A-V-~.)

is dense in LP if and

only

if

v 01. 5 =. 11 ?-~.11c1.

(9)

158 M. A PERELMUTER

3.

The analysis

of the

proof

shows that the restriction V- may be

replaced by

the weaker condition

1]

V_

]~_~= ~(s).

4.. The main device in the

proof

of the theorem 2 is the

inequality

N. Th.

Varopoulos [12]

has

proved

the abstract version of this

inequality

for the

generators

of submarkovian

semigroups.

Thus our result may be

generalized

for these

operators.

COROLLARY 3. - L~

Then ( - A + V) C~0

is

essentially self-adjoint

operator in

L 2.

Corollary 3

is

proved

under the restriction Ve but

using

the

results of C.

Simader [13]

or H.

Brezis [14],

we have

THEOREM 4. - Let

+

V) C~0

is

essentially self-adjoint.

Remark. - In the case 1&#x3E; 5 theorem 4 is a

generalization

of the Kalf- Walter-Schmincke-Simon theorem

(see [ 11 ], §

X.

4).

Example. -

We consider the

N-particle

Hamiltonian

Annales de l’Institut Henri Poincaré - Physique théorique

(10)

159 SCHRÖDINGER OPERATORS

Hardy inequality jJ= 2 N 03B1 (l - 2)2.

one sees

easily

that

V. ".

If 03B1 ~ (l-4) 2N

it

follows

from drorem 4 that H is essen-

tially self-adjoint 9cf. {15]. [16].

Note. that for thc fehtivisdc Hamiltonian

Lieb and

Thimng

haw

proved

that

with 03B2

= O

(N ex)

and is unbounded from below tor This correla-

tion

between 03B2

and N appears to be true for our nonrdativistic Hamil- tonian .as well so that the

dependence 03B1 = o(1/N)

is

optimaL

To prove

m-accretivity

we need.

THEOREM 5

1’1- -

Let

is accretive operator in Le.

We will show that

where

~v, w]

is semi-inner

product ~(~~ ~ ~ ~ ~, ~ X.8).

In the case of LP

spaces

so that

(18)

may be written in the

following

way

Vol. 52, n° 2-1990.

(11)

160 M. A. PERELMUTER

1.

The

sharpness

of the result is shown in

[6] by

the

example

2. T. Kato

~7~

raised the

problem

of

quasi-accretivity

of the

operator

-A+V in the

limiting

case

p= 1

for the

potentials

such that

We

will. show that the answer is

negative,

i. e., and if -A-V is

quasi-accretive

in

L 1,

then V E LOO.

Indeed,

the semi-inner

product

in

L

1

is

[~ u~~

= / 1"

~~;-) IJ will.

The

quasi-accretivity implies

that

B 117 I

Note

= 0, Therefore, (V,

i.

3. If

V E Lp, ,~ &#x3E; l/2, l &#x3E;_ 3,

then - A - V is the

generator

of the

positivity preserving Co-semigroup

in

L which

is not

quasi-contractive (other examples may

be found in

[18], [6]).

Theorems

2,

5 and

Lumer-Phillips

theorem

(see [ 11 ]., § X.8) imply

THEOREM 6. - Let

~.~here

k = (

-~-

a) ~ -1

+

.~/ 1- . ~

Then the closure

of

the o p era-

tor

( - A

+ V + C

(~)) r C~

is the generator

of Co-semigroup of

contractions in LP.

Remarks. - 1. Theorem 6 is

generalization (when

supp V is

compact)

of the result stated in

[4].

It is of interest to prove the LP-version of the result of C. Simader

[13]

and H.

Brezis [14].

2.

Having completed

this paper the author was informed

by

Yu. A.

Semenov about the

proof

of the theorem 6

provided

l’Institut henri Poincaré - Physique théorique

(12)

SCHRODINGER OPERATORS 161

REFERENCES

[1] A. Ya. POVZNER, On the Expansions of Arbitrary Functions in Terms of Eigenfunctions

of the

Operator -0394u+Cu,

Mat. Sbornik, Vol. 32, No. 1, 1953, pp. 109-156 ; A.M.S.

Trans. Ser. 2, Vol. 60, 1967, pp. 1-49.

[2] M. A. PERELMUTER, Positivity Preserving Operators and One Criterion of Essential

Self-Adjointness, J. Math. Anal. Appl., Vol. 82, No. 2, 1981, pp. 406-419.

[3] Yu. A. SEMENOV, One Criterion of Essential Self-Adjointness of the Schrödinger Operator with Negative Potential, Kiev, preprint, 1987.

[4] V. F. KOVALENKO, M. A. PERELMUTER, Yu. A. SEMENOV, Schrödinger Operators with

Ll/2w (Rl)-Potentials,

J. Math. Phys., Vol. 22, No. 5, 1981, pp. 1033-1044.

[5] Yu. A. SEMENOV, On the Spectral Theory of Second-Order Elliptic Differential Ope-

rators, Math. U.S.S.R. Sbornik, Vol. 56, No. 1, 1987, pp. 221-247.

[6] V. F. KOVALENKO and Yu. A. SEMENOV, Lp-Contractivity of the Semigroup Generated by the Schrödinger Operator with Singular Negative Potential, Kiev, preprint, 1985.

[7] T. KATO, Lp-theory of Schrödinger Operators with a Singular Potential, in Aspects of Positivity Funct. Anal. Proc. Conf., Tübingen, 24-28 June 1985, Amsterdam e. a., 1986,

pp. 41-48.

[8] W. P. NOVINGER, Mean Convergence in Lp Spaces, Proc. Am. Math. Soc., Vol. 34, No. 2, 1972, pp. 627-628.

[9] R. A. ADAMS, Sobolev Spaces, New York e. a., Academic Press, 1975.

[10] D. KINDERLEHRER and G. STAMPACCHIA, An Introduction to Variational Inequalities

and Their Applications, New York e. a., Academic Press, 1980.

[11] M. REED and B. SIMON, Methods of Modern Mathematical Physics, New York, San Francisco, London, Academic Press, 1978.

[12] N. Th. VAROPOULOS, Hardy-Littlewood Theory for Semi-Groups, J. Funct. Anal., Vol. 63, No. 2, 1985, pp. 240-260.

[13] C. G. SIMADER, Essential Self-Adjointness of Schrödinger Operators Bounded from Below, Math. Z., Vol. 159, No. 1, 1978, pp. 47-50.

[14] H. BREZIS, "Localized" Self-Adjointness of Schrödinder Operators, J. Oper. Theor., Vol. 1, No. 2, 1979, pp. 287-290.

[15] M. COMBESCURE-MOULIN and J. GINIBRE, Essential Self-Adjointness of Many-Particle Schrödinger Hamiltonians with Singular Twobody Potentials, Ann. Inst. H. Poincaré, Vol. A23, No. 3, 1975, p. 211-234.

[16] V. F. KOVALENKO and Yu. A. SEMENOV, Essential Self-Adjointness of Many-Particle

Hamiltonian Operators of Schrödinger Type with Singular Two-Particle Potentials,

Ann. Inst. H. Poincaré, Vol. A24, No. 4, 1977, pp. 325-334.

[17] E. H. LIEB and W. E. THIRRING, Gravitational Collapse in Quantum Mechanics with Relativistic Kinetic Energy, Ann. Phys. (U.S.A.), Vol. 155, No. 2, 1984, pp. 494-512.

[18] C. J. K. BATTY and E. B. DAVIES, Positive Semi-Groups and Resolvents, J. Oper.

Theor., Vol. 10, No. 2, 1983, pp. 357-363.

(Manuscript received March 23, 1989.)

Vol. 52, n° 2-1990.

Références

Documents relatifs

In sections 3 and 4 we use the p-adic Z-transform to give new proofs of several known results : the Mahler expansion with remainder for a continuous function, the

L’accès aux archives de la revue « Annales scientifiques de l’Université de Clermont- Ferrand 2 » implique l’accord avec les conditions générales d’utilisation (

Given two solutions u and v of the Schrödinger map Equation 1, we construct the two primitive solutions γu and γv of the binormal curvature flow equation whose derivatives coincide

By analogy to the classical theory then, this motivates the introduction of a family of L-adapted Hardy spaces HLp for all 0 < p < ∞, not equal to Lp in the range p ≤ p− L, on which

— En effet, les deux membres de la formule de représentation intégrale sont holomorphes dans Im k&gt;0, continus dans Im fe^O, ont les mêmes zéros avec même ordre dans Im k&gt;0 et

Mais rien ne prouve en général que l'opérateur A est autoadjoint et l'on a des exemples d'opérateurs pour lesquels A n'est pas autoadjoint et même que l'on ne peut étendre de façon

Observe also that estimate (4.5) shows only that the resolvent is bounded on [ 0, ∞[ and is not sufficient to apply the Hille-Yosida theorem and prove results for parabolic

We shall use complex interpolation to obtain item 1 of Theorem 1.3. This method is based on the boundedness of imaginary powers of H and of the Laplace-Beltrami operator. The