ccl intro lecture
TRANSCRIPT
-
7/30/2019 CCL Intro Lecture
1/39
1
An Introduction toComputational
Chemistry Laboratory
An Introduction toComputational
Chemistry Laboratory
-
7/30/2019 CCL Intro Lecture
2/39
2
What is Computational Chemistry
Laboratory (CCL)?
CCL is a virtual chemistry laboratory (in many cases
substitutes a real laboratory.)
The aim: use of computers to aid chemical inquiry. Based on:
Physical background
theory (Classical Newtonian or
Quantum Physics) Mathematical numerical algorithms (optimization, linearalgebra, iteration procedures, numerical integration etc.) Computer software and hardware (HYPERCHEM 8.0,GAUSSIAN03 on Windows PC)
Chemical knowledge and intuition for understanding and
interpretation of the computational results
-
7/30/2019 CCL Intro Lecture
3/39
3
Potential Energy Surface (PES) the main
chemistry inquiry
Chemistry is knowing the energy as a function of nuclear coordinates F. Jensen
-
7/30/2019 CCL Intro Lecture
4/39
4
Potential energy surfaces (and
similar properties) calculation
Classical (Molecular) Mechanics
quick, simple; accuracy depends on parameterization; no consideration of electrons interaction)
Quantum Mechanics:1.
Molecular Wave Function Theory
Ab initio molecular orbital methods...much more demandingcomputationally, generally more accurate.
Semi-empirical molecular orbital methods ...computationallyless demanding than ab initio, possible on a pc for moderatesized molecules, but generally less accurate than ab initio,especially for energies.
2.
Density functional theory more efficient and often moreaccurate than Wave Function based approaches.
-
7/30/2019 CCL Intro Lecture
5/39
5
Molecular Mechanics a theory
of molecules without electrons
Employs classical
(Newtonian) physics
Assumes Hookes Lawforces between atoms
(like a spring betweentwo masses)
EEstretchstretch
== kkss
(l(l --
lloo ))22
graph:graph: CC--CC;; C=OC=O
Force field = {kkss
}}
-
7/30/2019 CCL Intro Lecture
6/39
6
Molecular Mechanics
More elaborate Force Fields (FF)
-
7/30/2019 CCL Intro Lecture
7/39
7
Birth of quantum chemistry
Wave properties of matter
Birth of quantum chemistry
Wave properties of matter
Prince Louis de Broglie (1923):
=
h/mv= h/p
h= 6.62607550D-34 Js
p e
-i2x/
= e-i2px/h
(wave-particle duality
paradox)
-
probabilistic (statistic) wave.
Waves properties: interference, diffraction etc.
Possible explanation:The probabilistic behavior of elementary particles areconnected with structure of quantum vacuum.
-
7/30/2019 CCL Intro Lecture
8/39
8
Basis of Quantum ChemistryBasis of Quantum Chemistry
Postulate I : A closed system is fully described by
Postulate II: Operator
for every physical quantity
(-ih/2)d/dx (e-i2px/h
p (e-i2px/h
(-ih/2)d/dx
(p
p
(p
Schrdinger equation (1926):
(can be solved exactly for the Hydrogen atom, but nothing larger)
P.A.M. Dirac, 1929: The underlying physical laws necessary for themathematical theory of a large part of physics and the whole ofchemistry are thus completely known.
di H E
dt
-
7/30/2019 CCL Intro Lecture
9/39
9
One-dimensional Schrdinger wave
equation
2
2
2
22 2
2
2 2
2
2 2
2
Total energy kinetic potential
12 2
then
2
2
H E
pmv V V m
dp i p pp
dx
d d dp i i
dx dx dx
d
H Vm dx
dV E
m dx
Hamiltonian operator = operator of energy
SE = energy eigen-value equation
Extracts total energy, E
Many solutions
E
0
, E
1
,
E
n
x) wavefunction No direct physical meaning
x)|
Probability offinding particle withenergy E at point x
Single-valued, finite, continuous
-
7/30/2019 CCL Intro Lecture
10/39
10
Molecular
Schrdinger equation (SE):
1MA 2A_ h282H = 2a_ h282melectrons
A
>
nuclei
a
ZArAa_e2nuclei electrons
aA
ZAZB
rABe2
nuclei
A 1
rabe2
electrons
baB >++
kinetic energy (nuc.) kinetic energy (elect.)
2 kinetic energy terms plus3 Coulombic energy terms:
(one attractive, 2 repulsive)
H = E
H = Hamiltonion operator
-
7/30/2019 CCL Intro Lecture
11/39
11
Relativistic quantum mechanics
Dirac equation (1928) :
-
7/30/2019 CCL Intro Lecture
12/39
12
Influence of Relativity on
Quantum World and vice-versa
(x)
(x)( )
(x)
(x)
L
L
S
S
x
REALITY IS RELATIVISTIC AND THUS IS QUANTUM (and vice-versa!)
-
7/30/2019 CCL Intro Lecture
13/39
13
Diracs sea of electrons.
Quantum vacuum.
-
7/30/2019 CCL Intro Lecture
14/39
14
The NR molecular wavefunction
physical meaning
The wavefunction, , is a key quantity in quantum chemistry.
depends on coordinates and spins. Spin of electron
relativistic
property, additional discrete
coordinate ; |ms1
|=1/2
In a three dimensional system ofn-electrons,
is the probability of simultaneouslyfinding electron 1 with spin ms1
in the volume dx1
dy1
dz1
at (x1
,y1
,z1
),
electron 2 with spin
ms2
in the volume dx2
dy2
dz2
at (x2
,y2
,z2
)
and so on
The wave function should be normalized, that is, the probability offinding all electrons somewhere in space equals 1.
2
1 1 1 1 1,..., , ,..., ....n s sn n n nx z m m dx dy dz dx dy dz
2
1 1 1 1 1 1
... , , ,..., , , ... 1n n n n n nall m
x y z x y z dx dy dz dx dy dz
-
7/30/2019 CCL Intro Lecture
15/39
15
Wavefunctions general
properties
The wave function should be antisymmetric, that is, should
change sign when two electrons of the molecule interchange:
We can use the molecular wavefunction to calculate anyproperty of the molecular system. The average value, ,
of a
physical property of our molecular system is:
where, , is the quantum mechanical operator of the physicalproperty and
* C C d C
1 1 1
... ... n n nall m
dx dy dz dx dy dz d
1 1 1 ,..., 1
1 1 1 1
, , ,..., , , , , ,..., , , , ,...,
, , ,..., , , ,..., , , ,..., , , , ,...,
i i i j j j n n n s sn
j j j i i i n n n s sn
x y z x y z x y z x y z m m
x y z x y z x y z x y z m m
-
7/30/2019 CCL Intro Lecture
16/39
16
Ab-initioWavefunction approachAb-initioWavefunction approach
Simplifying assumptions are employed to solve the Schrdinger equation approximately:
Born-Oppenheimer approximation
allows separate
treatment of nuclei and electrons Hartree-Fock independent electron approximationallows each electron to be considered as beingaffected by the sum (field) of all other electrons.
LCAO Approximation
Tools: Variational
Principle or Perturbation
Theory
-
7/30/2019 CCL Intro Lecture
17/39
17
Born-Oppenheimer
Approximation -
1
MA2A
_ h282H
=
2a
_ h282m
electrons
A
>
nuclei
a ZArAa
_e2
nuclei electrons
aA
ZAZBrABe2nuclei
A 1rabe2
electrons
baB >
++
kinetic energy (nuc.) kinetic energy (elect.)
1 kinetic energy term plus2 Coulombic energy terms:(one attractive, 1 repulsive)
plus a constant for nuclei
constant
-
7/30/2019 CCL Intro Lecture
18/39
18
Steps of solution of the Schrdinger
equation in the Born-Oppenheimer
approximation:
Htot
= (Tn
+ Vn
)
+ Te
+ Vne
+ Ve
= (Hn
)
+He
1.
Electronic SE:Hee
(r,R)=Ee
(R) e
(r,R)
2.
Nuclear SE: (T
n + V
n + E
e (R) )
n (R)=E
n
n (R)
Vn
+ Ee
(R) = potential energy surface (PES)TOTAL WF
: (r,R) =n
(R) e
(r,R)
In our laboratory we concentrate mainly onsolution of the electronic SE and working with PES(finding minimums, transition states etc.)
-
7/30/2019 CCL Intro Lecture
19/39
19
Solving the Electronic SE:
Hartree-Fock (HF) approximation the physical background
Solving the Electronic SE:
Hartree-Fock (HF) approximation
the physical background
Multi-electronic SE:Hee
(r,R)=Ee
(R) e
(r,R)
is still very
complicated reduce it to the single-electronic equation
HF assumes that each electron experiences all the othersonly as a whole (field of charge) rather than individualelectron-electron interactions.
Instead of multielectronic Shrdinger equation introducesa one-electronic
Fock operatorF:
Fwhich is the sum of the kinetic energy of an electron, apotential that one electron would experience for a fixednucleus, and an average of the effects of the otherelectrons.
-
7/30/2019 CCL Intro Lecture
20/39
20
Mathematical foundation of the HF (or
Self-consistent-field (SCF)) method
Molecular orbital theory approximates the molecular wave function
as a antisymmetrized
product of orthonormal
one-electron
functions (or molecular spin-orbitals)
where is the antisymmetrization operator and
where k=1/2; 1/2 =
; -1/2
=
.
The antisymmetrization
operator is defined as the operator that
antisymmetrizes a product of n one-electron functions andmultiplies them by normalization factor(n!)-1/2
1 2( .... )nf f f
( , , )i i i i i k f x y z
-
7/30/2019 CCL Intro Lecture
21/39
21
Determinant of Slater
The antisymmetrized WF can berepresented as the Slaters determinant:
1 1 1 1 2 1 2 1 /2 1
1/2 1 2 1 2 2 2 2 2 /2 2
1 1 2 2 /2
( ) (1) ( ) (1) ( ) (1) ( ) (1) ... ( ) (1)( ) (2) ( ) (2) ( ) (2) ( ) (2) ... ( ) (2)
!... ...
( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ... ( ) ( )
n
n
n n n n n n
x x x x xx x x x x
n
x n x n x n x n x n
-
7/30/2019 CCL Intro Lecture
22/39
22
Variational Principle
The energyE
calculated from any
approximation
of the wavefunction
will be higher
than the true
energyE0
:
The better the wavefunction, the lower the energy(the more closely it approximates reality).
Changes (variation of parameters in ) are madesystematically to minimize the calculated energy.
At the energy minimum (which approximates thetrue energy of the system), dE/di
= 0.
*0
E H d E
-
7/30/2019 CCL Intro Lecture
23/39
23
The Hartree-Fock energy
functional
We shall restrict ourselves to closed shell configurations, for
such cases, a single Slater determinant is sufficient to describe
the molecular wave function. Using the variational
principle
within this framework lead to the restricted HF theory. The
Hartree-Fock energy for molecules with only closed shells is/ 2 / 2 / 2
1 1 1
12 (2 )
2
n n ncore
HF i ij ij
i i j
E H J K
21 1
1(1) (1) (1) (1) / (1)2
core core
i i i i I I i
I
H H Z r
12 12(1) (2)1/ (1) (2) , (1) (2)1/ (1) (2)
ij i j i j ij i j j i
J r K r
-
7/30/2019 CCL Intro Lecture
24/39
24
The Hartree-Fock equations
The Hartree-Fock equations are derived fromthe variational principle, which looks for thoseorbitals that minimizeEHF
.
For computational convenience the molecularorbitals are taken to be orthonormal:
The orthogonal Hartree-Fock molecular orbitals
satisfy the single-electronic equations:
(1) | (1)i j ij
(1) (1) (1)i i iF
-
7/30/2019 CCL Intro Lecture
25/39
25
The (Hartree-) Fock operator
Single-electronic operator:
The Coulomb operatorJj
and the exchange
operatorKj
are defined by
where
f is an arbitrary function
/ 2
21 1
1
1 (1) / 2 (1) (1)2
n
I I j j
I j
F Z r J K
2
2
12
*
2
12
1 (1) (1) (1) (2)
(2) (2) (1) (1) (1)
j j
jj j
f f dvr
fK f dvr
-
7/30/2019 CCL Intro Lecture
26/39
26
Next step:
MO-LCAO Approximation
Electron positions in molecular orbitals can be
approximated by a Linear Combination of AtomicOrbitals (LCAO).
This reduces the problem of finding the best
functional form for the molecular orbitals to themuch simpler one of optimizing a set of coefficients(cn
) in a linear equation:
= c1 c2 c3 c4 where
is the molecular orbital (MO) wavefunction
and
n
represent atomic orbital (AO) wavefunctions.
-
7/30/2019 CCL Intro Lecture
27/39
27
One step more:
Basis sets (BS)
A basis set
is a set of analytical functions (k
) used to
represent the shapes of atomic orbitals n:
General
contracted BS: n
=k bk(n)
k(n)
Contraction coefficients are calculated in a separate
atomic HF calculation;if k=1 basis set is called uncontracted.
Basis sets in common use have a simple mathematicalform for representing the radial distribution of electron
density.
Most commonly used are Gaussian basis sets, whichapproximate the better, but more numericallycomplicated Slater-Type orbitals (STO).
-
7/30/2019 CCL Intro Lecture
28/39
28
Hartree-Fock Self-ConsistentField (SCF) Method.
Hartree-Fock Self-Consistent
Field (SCF) Method.
Computational methodology (Jacobi iterations):
1. Guess the orbital occupation (position) of an electron(set of MO coefficients {cn
})
2. Calculate the potential each electron would experiencefrom all other electrons (Fock operatorF ({cn
}))
3.
Solve for Fock equations to generate a new, improved
guess at the positions of the electrons (new {cn })4. Repeat above two steps until the wavefunction for theelectrons is consistent with the field that it and the otherelectrons produce (SCF).
-
7/30/2019 CCL Intro Lecture
29/39
29
Types of HF
Multiplicity (M) = 2*S+1
(S is the total spin of the system)
Electrons can have spin up or down
. Most
calculations are closed shell calculations (M=1),using doubly occupied orbitals, holding two
electrons of opposite spins. RHF restricted HFOpen shell systems (M>1)
are calculated by
1.
ROHF
restricted open shell HF
the same
spatial orbitals for different spin-orbitals from thevalence pair;2.
UHF
unrestricted HF
different spatial parts for
different spins from the same valence pair
-
7/30/2019 CCL Intro Lecture
30/39
30
Illustrating an RHF singlet, and
ROHF and UHF doublet states
-
7/30/2019 CCL Intro Lecture
31/39
31
Semi-empirical MO Calculations:Further Simplifications of HF
Semi-empirical MO Calculations:
Further Simplifications of HF
Neglect core (1s) electrons; replace integral for Hcore
by an
empirical or calculated parameter
Neglect various other interactions between electrons on
adjacent atoms: CNDO, INDO, MINDO/3, MNDO, etc.
Add parameters so as to make the simplified calculation giveresults in agreement with observables (atomic spectra ormolecular properties).
/ 2 / 2 / 2
1 1 1
12 (2 )
2
n n ncore
F ii ij ij
i i j
E H J K
21 1
1(1) (1) (1) (1) / (1)2
core core
ii i i i I I i
I
H H Z r
12 12
(1) (2)1/ (1) (2) , (1) (2)1/ (1) (2)ij i j i j ij i j j i
J r K r
-
7/30/2019 CCL Intro Lecture
32/39
32
Beyond the SCF.
Correlated Methods (CM)
Include more explicit interaction of electrons than HF :
Ecorr
= E -
EHF , where E= H
Most CMs
begin with HF wavefunction, then incorporate
varying amounts of electron-electron interaction by mixing in
excited state determinants with ground state HF determinant
The limit of infinite basis set
& complete electron correlation
is
the exact solution of Schrdinger equation (which is still an
approximation)
-
7/30/2019 CCL Intro Lecture
33/39
33
Beyond the SCF.
Correlation effects on properties.
-
7/30/2019 CCL Intro Lecture
34/39
34
Two alternative ways of the
electron correlation treatmentHF (Hartree-Fock)
a singe determinant
theory
no correlation included!1. WF based multi-determinant
correlation methods:
1.
Configuration Interaction
(CI)
Variational; CISD, CSID(T)
2.
Many-body perturbation theory
(including coupled cluster)
Non-variational; MBPT2, MBPT3; CCSD; CCSD(T)2. Density functional theory (DFT)
correlation method not based on
wave-function, but rather on modification of the energy functional:
A single determinant theory including correlation!/ 2 / 2 / 2
1 1 1
2 ( )n n n
core Exch Corr
DFT i ij ij
i i j
E H J X
-
7/30/2019 CCL Intro Lecture
35/39
35
Summary of Choices:
6-311++G**
6-311G**
6-311G*
3-21G*
3-21G
STO-3G
STO
HF CI QCISD QCISDT MP2 MP3 MP4 correlation
increasingsize of
basis set
increasing level of theory
Schrodinger
increasing accuracy,increasing cpu time
infinite basis set
full e
-
-
7/30/2019 CCL Intro Lecture
36/39
36
The extra dimension:
-
7/30/2019 CCL Intro Lecture
37/39
37
Summary: Levels ofQMTheory
H=E
Born-Oppenheimer approximation
Single determinant SCF
Semi-empirical methods
Correlation approaches:
1.Multi-determinantial (MCSCF, CI, CC, MBPT)
2. Single determinantial
(DFT)
-
7/30/2019 CCL Intro Lecture
38/39
38
Some applications during your
work...
Calculation of reaction pathways (mechanisms)
Determination of reaction intermediates andtransition structures
Visualization of orbital interactions (formation of newbonds, breaking bonds as a reaction proceeds)
Shapes of molecules including their chargedistribution (electron density)
NMR chemical shift prediction.
IR spectra calculation and interpretation.
-
7/30/2019 CCL Intro Lecture
39/39
39
Method Type Features Advantages Disadvantages Best for
Molecular
Mechanic
Uses classical
physics Relies on force-field
with embeddedempirical parameters
Computationallyleast intensive - fast
and useful withlimited computerresources
Good for:
Enthalpy ofFormation(sometimes)
Dipole Moment
Geometry (bondlengths, bond angles,
dihedral angles) oflowest energyconformation.
Particular force
field, applicable onlyfor a limited class ofmolecules
Does not calculateelectronic properties
Requires
experimental data(or data from ab
initio
calculations)
Large systems
(~1000 of atoms) Can be used for
molecules as large asenzymes
Systems or processeswith no breaking or
forming of bonds
Semi-Empirical Uses quantum
physics Uses experimentally
derived empiricalparameters
Uses manyapproximations
Less demandingcomputationally thanab
initio
methods
Capable ofcalculating transitionstates and excitedstates
Requiresexperimental data(or data from ab
initio) forparameters
Less rigorous thanab
initio) methods
Medium-sizedsystems (hundreds ofatoms)
Systems involvingelectronic transition
Ab
Initio Uses quantum
physics Mathematically
rigorous, noempirical parameters
Uses approximationextensively
Useful for a broadrange of systems
does not depend onexperimental data
Capable ofcalculating transitionstates and excitedstates
Computationallyexpensive
Small systems (tensof atoms)
Systems involvingelectronic transition
Molecules withoutavailableexperimental data
Systems requiringrigorous accuracy