Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4...

36
Roi Baer Institute of Chemistry, The Fritz Haber Center for Molecular Dynamics The Hebrew University of Jerusalem, Jerusalem, Israel

Transcript of Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4...

Page 1: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

Roi BaerInstitute of Chemistry,

The Fritz Haber Center for Molecular Dynamics

The Hebrew University of Jerusalem, 

Jerusalem, Israel

Page 2: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

• Ms. Ester Livshits (PhD student, Jerusalem)• Dr. Helen Eisenberg (Postdoc, Jerusalem)• Tamar Gershon (PhD student Jerusalem)Collaborators • Tamar Gershon (PhD student, Jerusalem)• Professor Leeor Kronik (Weizmann)• Professor Daniel Neuhauser (UCLA)• Professor Anna Krylov (USC)

Collaborators and Students

• Professor Anna Krylov (USC)

• Israel Science Foundation (ISF)• US-Israel Binational Science Foundation

(BSF)Funding (BSF)• FAME network of excellence

Funding

Page 3: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

• An electron is repelled not only by other electrons, but also by itselfSelf

• Good at the short range (emulates correlation)

• But much too much at long range 

Self repulsion

• Static polarizability• Stability of anions• Charge quantization in decoupled systems

Unphysical b h i

g q p y• Charge transfer excitations• Rydberg excitations

behavior

Page 4: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

1 1 2 2Take a charge at and bring another charge from infinity to . q qr r

1 22 12 1 2

12

The energy for this i .s: q q

E rr

= = −r r

1 3 2 3Next bring another charge from infinity to :q q q q

q E = +r3 3 313 23

Next, bring another charge from infinity to :q Er r

= +r

1The total energy is a sum over pairs:

2i j

i j ij

q qE

r≠

= ∑2 i j ijr≠

( ) ( )( )3 31

For continuous charge dist: ( density of charge)2

n nE d rd r n

′′=

′−∫r r

rr r

But this includes self interaction - analogous to the termsi j=

( ) ( ) ( ) ( ) ( )In Hartree-Fock theory we have: , ; ,occN

i inρ φ φ ρ′ ′= =∑r r r r r r r( ) ( ) ( ) ( ) ( )

( ) ( ) ( )1

2

3 3 3 3r,r1 1

The 2-electron repulsion is: d rd r2 2 r-r

i

ee

n nE d rd r

ρ=

′′′ ′= −

′ ′−∫ ∫r r

r rDirect or Hatree Exchange

So in HF theory self-repulsion is automatically removed.

Page 5: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

.5

V(x) .3

.4

High and widebarrier

V0 0

.1

.2

x-2.0 -1.5 -1.0 -.5 0.0 .5 1.0 1.5 2.0

0.0

Self repulsionNo self repulsion

.3

.4

.5

High and widebarrier

p

.3

.4

.5

High and widebarrier

V(x)

.1

.2

.3 ba e

V(x)

.1

.2

x-2.0 -1.5 -1.0 -.5 0.0 .5 1.0 1.5 2.0

0.0

x-2.0 -1.5 -1.0 -.5 0.0 .5 1.0 1.5 2.0

0.0

Page 6: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

• LSDA/GGA: no barrier• Scheffler et al:

– Electron to O2 too early – O - is strongly attracted– O2 is strongly attracted– Introduced “spin rules”– Lowering the spin on the

triplet O2 was forbidden – This is a strange ad hoc

mechanism

J. Behler, B. Delley, S. Lorenz, K. Reuter, and M. Scheffler,, Phys. Rev. Lett. 94 (3), 036104 (2005).

Page 7: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

E. 

PBE

Livshits, R

B3LYP

R. Baer andB3LYP d R. Koslof

BNL

ff, work in BNL progress ((2009)

Page 8: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

1 1 121 1ˆ2

ˆ2

1j

j j j jj

Ur

T′≠ ′

∇ == − ∑ ∑

( )Unique gs WF of electrons having gs density nψ = r( )q g g g y gs

ψ

( )0Unique gs WF of non-int Fermions having gs density nψ = r

ˆ ˆˆgs gs

E T V Uψ ψ= + +

0 0ˆ ˆˆ

CT V U E nψ ψ ⎡ ⎤= + + + ⎢ ⎥⎣ ⎦

( ) ( ) 30 0

ˆ ˆgs gsV v n d r Vψ ψ ψ ψ= =∫ r r

Page 9: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

ˆ ˆˆ ˆ⎡ ⎤0 0C gs gs

E n T U T Uψ ψ ψ ψ⎡ ⎤ = + − +⎢ ⎥⎣ ⎦

0 0ˆ ˆ

C gs gsT n T Tψ ψ ψ ψ⎡ ⎤ = −⎢ ⎥⎣ ⎦

⎡ ⎤ ⎡ ⎤0 0C CE n T n⎡ ⎤ ⎡ ⎤< >⎢ ⎥ ⎢ ⎥⎣ ⎦ ⎣ ⎦

•Theorem: A physical quantity cannot have self-repulsion•Corollary: Expectation values of physical operators are self-repulsion free•Corollary: Ec and Tc are self repulsion-free

Page 10: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

U E n E nψ ψ ⎡ ⎤ ⎡ ⎤= +⎢ ⎥ ⎢ ⎥⎣ ⎦ ⎣ ⎦( )

0 0

Self-repulsion freeH X

U E n E nψ ψ +⎢ ⎥ ⎢ ⎥⎣ ⎦ ⎣ ⎦

( ) ( ) 3 31 n nE n d rd r

′⎡ ⎤ ′=⎢ ⎥⎣ ⎦ ∫∫

r r

( )2

Has self-repulsion

HE n d rd r=⎢ ⎥⎣ ⎦ ′−∫∫ r r

( )2

( )Has self-repulsion

( )3 3

,12X

E n d rd rρ ′

⎡ ⎤ ′= −⎢ ⎥⎣ ⎦ ′−∫∫r r

r r

( )Has self-attraction

Page 11: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

0 0ˆ ˆˆ ˆ

C gs gsE n T U T Uψ ψ ψ ψ⎡ ⎤ = + − +⎢ ⎥⎣ ⎦( )self repulsion-free

( ) ( )( )⎡ ⎤ ∫ ( ) ( )( )( )

3

ˆˆ ˆ

,SLC XC X

T U T E

E n f n d r En

ψ ψ ψ ψ+ +

⎡ ⎤ = −⎢ ⎥⎣ ⎦ ∇∫ r r

( )0 0gs gs HT U T Eψ ψ ψ ψ+ − +

Long range self repulsion

Page 12: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

0 0ˆ ˆˆ ˆ

C gs gsE n T U T Uψ ψ ψ ψ⎡ ⎤ = + − +⎢ ⎥⎣ ⎦ 0 0C gs gs

ψ ψ ψ ψ⎢ ⎥⎣ ⎦

( )0 0

1ˆ ˆ ˆC j jE n Y Y Y yψ ψ ψ ψ⎡ ⎤ = − = −⎢ ⎥⎣ ⎦ ∑ r r( )0 0 2C gs gs j j

jj

yγ γ γψ ψ ψ ψ′≠

⎢ ⎥⎣ ⎦ ∑

re γ− ( )erfc rγ1 1

0.25

0.30

ey

rγ =( )erfc r

yrγ

γ=

1 11

yrγ γ

=+

( ) ( )1

0 10

0.15

0.20

ˆ ˆ0 Y Y E T Eψ ψ ψ ψ

( ) ( )0

10y r y r

r ∞= =

2 3 4 5 6 7 8

0.05

0.10

0 00 :

gs gs C C CY Y E T Eγ γγ ψ ψ ψ ψ= − = − <

ˆ ˆ: 0Y Y Eγ ψ ψ ψ ψ= ∞ = >

Conclusion: There exists a for which there is equality γ0 0

: 0gs gs CY Y Eγ γγ ψ ψ ψ ψ= ∞ − = >

Page 13: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

0 0ˆ ˆ

C gs gsE n Y Yγ γψ ψ ψ ψ⎡ ⎤ = −⎢ ⎥⎣ ⎦ 0 0

ˆC gs gs

y y

gs gs H XY E Eγ γ

γ γ

γψ ψ⎢ ⎥⎣ ⎦

⎡ ⎤= − −⎢ ⎥⎣ ⎦Fix γ and findGood approx for f(n,dn):

( ) ( )( ) 3 yH b

•Becke (B3LYP)•…•Savin•Hirao( ) ( )( ) 3,

y

XC XHybC

f n nE d r E γγ ∇= −∫ r r

( ) ( )1∫∫

Hirao•Handy•Baer+Neuhauser•Yang

( )

( ) ( ) 3 31,

2y

XE y d rd rγ

γρ ′ ′ ′= − −∫∫ r r r r •Tozer•Truhlar•Head-Gordon•Gerber-Angyan

( )erfc1 1, ,

1

rr ey

r r r

γ

γ

γ

γ

=+

Gerber Angyan•Perdew-Scuseria•…

1 r r rγ+“B3LYP” typeLong-range self-repulsion

“RSH” typeNo long-range self-repulsion

Page 14: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

1 Cε

11

C

C Ctγ ε

=+ −

1.0g

1 3

1sr

n∝

0 6

0.8

0.4

0.6

0.2

0 5 10 15 20rs

Page 15: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

γ HOMOε− HOMOε−SCFΔγ 10a−wγ 10a−w

100

10

Unpolarized Polarized

1g

0.1

0.010 01 0 1 1 100.01 0.1 1 10

rs

E. Livshits and R. Baer, PCCP  9, 2932(2007)

Pair correlation function taken from:P. Gori-Giorgi and J. P. Perdew, Phys. Rev. B 66, 165118 (2002).

Page 16: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

• In B3LYP – OK• In RSH – NOT OK• Must be: γ system dependent

Page 17: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

30

25

HOMO B3LYP

Exp. Vertical Theorem:Theorem:

Th i i tiTh i i ti

15

20

(eV)

Basis: cc-pVTZThe ionization The ionization

potential of the potential of the

t i tlt i tl

10

IP system is exactly system is exactly

equal to the HOMO equal to the HOMO

5

energyenergy

0

• M. Levy and J. P. Perdew, in Density Functional Methods in Physics, edited by R. Dreizler and J. Perovidencia (Plenum, New York, 1985), pp. 11.

• C.‐O. Almbladh and U. von‐Barth, Phys. Rev. B 31, 3231 (1985).

Page 18: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

γ HOMOε− HOMOε−SCFΔγ 10a−wγ 10a−w

Molecule

γ(a0

-1)

IP exp

IPw=0

IPw=0.1

IPB3LYP

( );HOMO Nε γ

Li2 0.3 5.1 -0.1 0.0 -0.2N2 0.6 15.6 -1.0 -0.6 0.0O2 0.7 12.1 -1.2 -0.9 -0.7F 0 7 15 7 -0 8 -0 4 0 1

( )( ) ( ); 1;E N E Nγ γ= − −

F2 0.7 15.7 -0.8 -0.4 0.1Na2 0.4 4.9 -0.2 -0.1 -0.3P2 0.8 10.6 -0.4 -0.3 0.0S2 0.5 9.4 -0.4 -0.2 -0.2Cl2 0.5 11.5 -0.3 0.0 0.1BeH 0.6 8.2 -0.2 -0.1 -0.2CO 0.6 14.0 -0.3 -0.1 0.0HF 0 7 16 0 -0 4 0 0 0 0HF 0.7 16.0 -0.4 0.0 0.0HCl 0.5 12.8 0.0 0.3 0.2NH3 0.5 10.2 -0.6 -0.4 -0.5PH3 0.4 9.9 -0.6 -0.4 -0.6

E. Livshits and R. Baer, PCCP 9, 2932 (2007)

Mean -0.5 -0.2 -0.2

RMS 0.6 0.4 0.3

Page 19: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

‐0.48

‐0.52

‐0.50

HF

H2+

QCHEM 3.1

0 56

‐0.54

E(E h)

B3LYP

BLYP

cc-pVTZ

Same

‐0.58

‐0.56 Same problems appear in:

‐0.62

‐0.60 He2+, Ne2

+

etc0 1 2 3 4 5 6 7 8 9

H‐H distance (A)

R Merkle A Savin and H Preuss J Chem Phys 97 9216 (1992)R. Merkle, A. Savin, and H. Preuss, J. Chem. Phys. 97, 9216 (1992).Y. Zhang and W. Yang, J. Chem. Phys. 109, 2604 (1998).A. Ruzsinszky, J. P. Perdew, G. I. Csonka, O. A. Vydrov, and G. E. Scuseria, J. Chem. Phys. 126, 104102 (2007).

Page 20: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

LR R+

r → ∞

L

( )( ) ( ) ( )

1

2R R

GS L R

E r E E +

= ±

→ +

RR+

( ) ( ) ( )

4

R R

...2

correctE r E E

r

α

→ +

− +R

0 5 0 5( )1

Density Picture:

n n n+R0.5 R0.5( )

( ) ( )0.5

214

2

L Rn n n

E r E Rr

= +

= +

Page 21: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

‐0.50

‐0.48

By using RSH we get id f l i

0 54

‐0.52 HF

B3LYP

BLYP

rid of repulsion

But this is not enough

h l l‐0.56

‐0.54

E(E h) BLYP

γ=0.5There is also large error at r = r0Wrong as mptotic

‐0.60

‐0.58 Wrong asymptotic curve (≠ -α/2r4)

‐0.62

0 1 2 3 4 5 6 7 8 9

H‐H distance (A)

E. Livshits and R. Baer, J. Phys. Chem A 112, 12789 (2008).

Page 22: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

( ) 0 1 0.5 0.5E E Eγ γγ − −Δ = −

1 4

1.6

Ne2+

( ) γ γ

E HF γ→ ∞

0 8

1.0

1.2

1.4

)

Ne2+

He2+

H2+ ΔE

ΔE

0 2

0.4

0.6

0.8

ΔE (eV)

LDA γ → 0

ΔE

0 4

‐0.2

0.0

0.2

( )Åeqr

‐0.4

0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

γ (1/a0)

( )eq

2ωR+R+ R++RR0.5++R0.5+

E. Livshits and R. Baer, J. Phys. Chem A 112, 12789 (2008).

Page 23: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

E. Livs

• Have γ so: localized-delocalized degenerate• Have γ so: localized-delocalized degenerate

hits and R.

• Energies in kcal/mole• Energies in kcal/moleProperty R BLYP B3LYP HF LC LC* Exp

Baer, J. P

h

Property R BLYP B3LYP HF BNL BNL* Exp.

Enthalpy by atomization

H 66 65 60.9 60.9 60.9 61He 82 75 43 74 59 55N 75 60 2 59 34 32

hys. Chem

ANe 75 60 2 59 34 32

Enthalpy by asymptote

H NA NA 60.9 50 60.9 61He NA NA 43 42 59 55Ne NA NA 2 27 34 32H 1 1 1 1 1 06 1 2 1 06 1 05

A 112, 1278

H 1.1 1.1 1.06 1.2 1.06 1.05He 1.2 1.1 1.075 1.2 1.078 1.080Ne 1.9 1.9 1.7 1.760 1.72 1.765H 2.7 2.9 3.3 2.9 3.3 3.32H 1 7 2 0 2 5 2 1 2 5 2 42

( )eqr Å

89 (2008).

He 1.7 2.0 2.5 2.1 2.5 2.42Ne 0.5 0.6 0.9 0.726 0.8 0.729H NA NA 1 0.6 1 1He NA NA 0.98 NA 0.98 1N NA NA 1 01 NA 1 02 1

R effα α ( ) 51

2effE r rα ′=

R refα α

Ne NA NA 1.01 NA 1.02 1H 5.3 5.6 4.51 5.8 4.51 4.50He 1.6 1.5 1.34 1.8 1.41 1.38Ne 3.1 2.9 2.4 3.2 2.70 2.66

ff

( )30

daugRa a

( )2ff

Page 24: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

By using γ* several y g γproblems are solvedp

Repulsion instead of

i

No degeneracy

b kiProper well depth at r=r0

Fine bond lengths

“Good vibrations”

Leading term in asymptotic

form of i l

Excellent atomic l i biliattraction breaking depth at r r0 lengths vibrations potential -

correctpolarizability

E. Livshits and R. Baer, J .Phys. Chem, A 112, 12789 (2008).

Page 25: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

Energies (kcal/mol) of (H2O)2+ relative to 

Method Vertical neutral

Hemi-bonded

2 2the proton‐transferred geometry

neutral bonded

BLYP 6.7 ‐8.3B3LYP -1.6

E. Livshits, A. I. Krylov and R. Baer, work in progress 

BNL*(0.56) 23.3 8.2EOM-IP-CCSD 20.0 5.3EOM-IP-CC(2.3) 21.8 7.4

p g(2009).

Neutral dimer Hemi‐bonded

Proton‐transferred

Page 26: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

• We looked into the Anthracene-TCNE charge transfer complex in solution (CH2Cl2)

1• In ideal CT excitation: 1CT ArH A

h IP EAR

νε

= − −

J. M. Masnovi, E. A. Seddon, and J. K. Kochi, Can. J. of Chem . 62 (11), 2552 (1984).

Page 27: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

• First stage is to find the appropriate γ• This requires finding a γ that gives proper q g γ g p p

HOMO energies for IP(ArH) and EA(TCNE)γ IP (ANT) (eV) IP (TCNE-) = EA (eV) Σrγ IP (ANT) (eV) IP (TCNE ) = EA (eV) Σr

ΔESCF ‐εHOMO r ΔESCF ‐εHOMO r0.01 6.09 4.03 0.34 0.59 -0.90 2.53 2.870 02 6 09 4 22 0 31 1 69 0 73 1 43 1 740.02 6.09 4.22 0.31 1.69 -0.73 1.43 1.74

0.033 6.14 4.44 0.28 1.72 -0.52 1.30 1.580.05 6.18 4.71 0.24 1.77 -0.24 1.14 1.38

0 3 6 91 7 02 0 02 2 44 2 34 0 04 0 060.3 6.91 7.02 -0.02 2.44 2.34 0.04 0.060.4 7.04 7.35 -0.04 2.69 2.88 -0.07 0.120.5 7.08 7.56 -0.07 2.89 3.27 -0.13 0.20

• One obtains an optimal value of 0.3

Page 28: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

• Checking the Mulliken law in gas phase3.8

3 6

3.7TDDFT g=0.3

Mulliken

3.5

3.6

E (eV)

3 3

3.4

3.2

3.3

0 02 0 025 0 03 0 035 0 04 0 045 0 050.02 0.025 0.03 0.035 0.04 0.045 0.05

1/R(a0)1CT ArH A

h I EAR

ν = − −

Page 29: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

• We took a B3LYP optimized geometry for the complex and performed TDDFT calculation

Ar B3LYP BNLγ=0.5

BNL γ* ExpΕ f γ* Ε f Ε f2 1 0 03 4 4 0 33 3 8 0 03 3 59 0 02Benzene 2.1 0.03 4.4 0.33 3.8 0.03 3.59 0.02

Toluene 1.8 0.04 4.0 0.32 3.4 0.03 3.36 0.03o‐Xylene 1.5 ~0 3.7 0.31 3.0 0.01 3.15 0.05Naphthalene 0.9 ~0 3.3 0.32 2.7 ~0 2.60 0.01

T. Stein, L. Kronik and R. Baer, JACS to be published (2008).

Page 30: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

Substituent PBE B3‐LYP

BNL Exp26Ε(γ=0.5)  γ∗ Ε(γ=0.3)

None 0.9 1.0 2.3 0.31 1.82 1.739‐cyano Fail 0.5 2.6 0.30 2.03 2.019‐chloro 0.9 1.0 2.3 0.31 1.82 1.749‐carbo‐ 0 8 0 9 2 4 0 30 1.84 1.849 carbomethoxy

0.8 0.9 2.4 0.30 1.84 1.84

9‐methyl 1.0 1.1 2.1 0.30 1.71 1.559‐nitro 0 6 0 9 2 8 0 30 2 12 2 039‐nitro 0.6 0.9 2.8 0.30 2.12 2.039.10‐dimethyl

1.3 1.4 2.1 0.30 1.77 1.44

9 formyl 0 8 1 0 2 5 0 30 1 95 1 909‐formyl 0.8 1.0 2.5 0.30 1.95 1.909‐formyl 10‐chloro

0.8 0.9 2.5 0.30 1.96 1.96

T. Stein, L. Kronik and R. Baer, JACS to be published (2008).

Page 31: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

• Kohn-Sham bad gaps (50-70% too small)• Because missing “derivative discontinuity”g y• Hartree-Fock theory terrible: 500% too large.

Page 32: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

LDAE E

γ−

g g

Hartree Fock LDA

g g

E E

E E−

Page 33: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

0.2

0.16

0.18 NaCl* C

ε=

0 1

0.12

0.14

∗ MgO

optAε −

0.22C =

0 06

0.08

0.1γ∗

AlPC

0.8A=

0.02

0.04

0.06 AlP

AlN

SiC AlA

GaP

Si

0

0 2 4 6 8 10 12 14

SiC AlAsAlSb

εOpt

Page 34: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

3.0

2.5

2.0

ap (e

V)

ExpGaP AlP

1.0

1.5

Ban

d G

a ExpgHSELSDA

AlSbSiC

0.5

1.0 LSDAAlSb

0.01 0 1 5 2 0 2 5

Si

1.0 1.5 2.0 2.5Experimental Band gap (eV)

H. Eisenberg, R. Baer, work in progress (2008).

Page 35: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

8.0

7.0

)

6.0

Ban

d G

ap (e

V)

Exp

γ

HSEMgO

4.0

5.0

B HSE

LSDA

AlN

MgO

NaCl

3.0

4.0

4 0 5 0 6 0 7 0 8 0

CNaCl

AlN

4.0 5.0 6.0 7.0 8.0

Experimental Band gap (eV)

H. Eisenberg, R. Baer, work in progress (2008).

Page 36: Roi Baer - Home | Institute for Mathematics and its … 1.6 1.5 1.34 1.8 1.41 1.38 Ne 3.1 2.9 2.4 3.2 2.70 2.66 (3) 0 daug R aa ff Byyg using γ* several problems are solved Repulsion

Dogmatic spirit Pragmatic spirit spirit

Hybrid (B3LYP/RSH) theory

γ system depende

nt

Ab initio ways to

determine γ

“Systems too difficult for DFT”

In B3LYP: In RSH: γγ

insensitive to density (factor 3)

γvery

sensitive (factor ~100)

Symmetric cation

radicals

Charge-transfer

excitationsBand gaps