5.

62 Spring 2004 Transition-State Theory

Lecture #34, Page 1

A good theory must take into account the internal degrees of freedom of the reactants and their angle of approach. An approach known as transition state theory (=activated complex theory)(=absolute rate theory) does so in an approximate way. POTENTIAL ENERGY SURFACE A correct theory must consider the internal structure of molecules and the forces acting on atoms in the molecules because bonds are being broken and formed during a reaction. During a reactive collision, the force on an atom depends on both intramolecular forces (forces between atoms in a molecule) and intermolecular forces (forces between molecules). Must treat the two colliding reactants as a single quantum mechanical system. This system exists only during collision process. The systems potential energy is calculated the same way the potential energy for nuclear vibrational motion is calculated. Within the BornOppenheimer approximation, solve

H el el = Eel el
at fixed nuclear configuration. The resulting Eel is the potential energy at that nuclear configuration. Systematically vary the nuclear configuration (grid of points) to get potential energy as a function of nuclear coordinates. Problem: too many nuclear coordinates! Cant plot potential energy as a function of more than 2 coordinates. A plot of potential energy versus 2 coordinates is a 3D plot where the potential energy is a SURFACE. Potential energy for more than 2 coordinates is still called a surface even through there are more than 2 coordinates.

5.62 Spring 2004 EXAMPLE: H2+FHF+H

Lecture #34, Page 2

For collinear approach: there are 2 independent variables RHF and RHH on which the potential depends

The potential energy of this system can be represented as a 3D surface as a function of RHF and RHH. The 3D surface an be represented as a 2D contour potential energy surface.

Lines on a contour map represent EQUIPOTENTIALS. Valleys correspond to initial and final states.

5.62 Spring 2004

Lecture #34, Page 3

(To go from one valley to another requires crossing a saddle point)

REACTION COORDINATE FOR F + H2

REACTION COORDINATE minimum energy path along deepest part of potential energy surface Note: reaction coordinate here corresponds to antisymmetric H2F vibration TRANSITION STATE transitory

structure dissociates within one half antisymmetric vibration TRANSITION STATE THEORY An approach to calculating a rate constant by reducing the dynamics of the reaction to an equilibrium between the reactants and the transition state along the reaction coordinate.

[H 2 F]

complex with a definite

A+B [AB] products

5.62 Spring 2004

Lecture #34, Page 4

Uses statistical mechanics to treat the equilibrium. The reaction coordinate is the 1D antisymmetric vibration of the transition state.

5.62 Spring 2004 Transition State Theory

Lecture #34, Page 5

Transition state theory = Activated Complex Theory = Absolute Rate Theory

Assume equilibrium between reactants H2+F and the transition state. Treat the transition state as a molecule with structure which unimolecularly decays with rate constant k.

[H 2 F] K = [H 2 ][F]

d [HF] = k[H 2 F] = kK [H 2 ][F] dt

k has units of sec-1 (unimolecularly decay). The motion along the reaction coordinate looks like an antisymmetric vibration of of this vibration. Therefore k can be approximated by the frequency of the antisymmetric vibration [sec-1] k= = frequency of antisymmetric vibration (bond formation and cleavage looks like antisymmetric vibration)
d [HF] = K [H 2 ][F] dt E / k T (q * / N ) d [HF] B e [H 2 ][F] = H* F 2 dt ( q / N )( q / N )

[H 2 F]

, one-half cycle

* qvib qrot g 0 (qtrans / N) K = H2 H2* H2 H2 F F (qtrans / N )(qtrans / N ) q vib q rot g 0 g 0

E / kBT e
[H 2 F]

Reaction coordinate is antisymmetric vibrational mode of

. This

vibration is fully excited because it leads to the cleavage of the H-H bond and the formation of the H-F bond. For a fully excited vibration,

5.62 Spring 2004

Lecture #34, Page 6

h << kBT The vibrational partition function for the antisymmetric mode is
*,asym = qvib

1 1 e h / kBT

k BT h

, since

e h / kBT 1

h k BT

*' qvib qrot g 0 (qtrans / N) K = H2 H2* H2 H2 F F (qtrans / N )(qtrans / N ) q vib q rot g 0 g 0

E / kBT e

where

*' vib

3 n 51

*' qvib =

i =1 3 n 61

1 e

1
h i / k BT

if the transition state is linear if the transition state is linear

1 e
i =1

1
h i / k BT

(n = # atoms in transition state)

is the partition function from which the antisymmetric vibrational mode

is excluded; it because the reaction coordinate So

K =

k BT ' K h

'

is the special

excluding the

partition function for the reaction coordinate

What is

5.62 Spring 2004

Lecture #34, Page 7

Since a molecule cannot have a vibrational energy lower than its zero point energy, the effective barrier along the reaction coordinate is

E = V0 + ( ZPE )TS ( ZPE ) R

H2F

, n=3, so 3n-5-1 = 3 vibrational modes, thus

E = V0 +

h 2 + bend H 2 sym.st . 2   

difference in ZPE

FORMULATION OF kTST

k TST =

k BT ' K h

but not all reactant molecules make it all the way to products some are reflected back to separated reactants. Thus,

k TST =

k BT ' K , h

where

= transmission coefficient

EVALUATION OF kTST POTENTIAL ENERGY SURFACE KNOWN:

E I

- directly from potential energy surface - calculate from geometric structure of transition state - analyze shape of potential in saddle point region - trajectory calculations (consider = 1 for now) T=300K, mH2 = 2 amu, mF = 19 amu

H2+F HF + H Translational part

5.62 Spring 2004

Lecture #34, Page 8

3/ 2

6 1023 0.021 1 23 = 6 10 mol 3/ 2 ( 2 1.38 1023 J / K 300K ) 0.002 0.019kg = 2.52 107 m3 / mol
Rotational part

( 6.63 1034 J s )

3/ 2

H = 2
2

=1

I H 2 = 4.56 1048 m 2 kg

8 2 Ik BT qrot = h2 I = 1.24 1046 m 2 kg

qrot I H2 = = 54.4 H2 qrot I H2

3n-5-1 = 3 vibrational degrees of freedom (one vibration is reaction coordinate)

[H 2 F]

is a linear transition state

h s = 5771K ( stretch) kB

h b = 573K (2 x degen bend ) kB

h H 2 kB

= 6323K

5.62 Spring 2004

Lecture #34, Page 9

1 e qvib = *H 2 qvib
*'

h s / k BT

(1 e

) (1 e
1
h H / k BT 2

h b / k BT

= 1.38

Electronic part

g0 = 4
Calculate E

F g0 =4

H2 g0 =1

g0 =1 F H2 g0 g0

V0 = 3.8kJ / mol
H = 1.32 1014 s 1
2

s = 1.20 1014 s 1

b = 1.19 1013 s 1 (How do we guess values for s and b ?)

1 E = V0 + hN s + 2 b H2 = 6.1kJ / mol 2
Calculate kBT/h

k BT 1.38 1023 J / K 300 K = = 6.24 1012 s 1 34 h 6.63 10 J s

k TST =

k BT ' K h

1 = 1( 6.24 1012 s 1 )( 2.52 107 m3mol 1 ) ( 54.4 )(1.38 ) e 6.1/ RT 3 = 3.93 107 e 6.1/ RT

k TST = 3.40 106 m3 mol 1s 1 at 300K

5.62 Spring 2004

Lecture #34, Page 10 - NOT BAD AGREEMENT!

k EXP = 2.70 106 m3 mol 1s 1

Experimental value is smaller because is probably not 1. Sometimes kTST will be smaller than kEXP because of tunneling. This model for kTST does not take the quantum mechanical phenomenon of tunneling into account. Tunneling can make the reaction rate become faster than the kTST prediction. If kTST < kEXP, it may mean that there is some tunneling contribution. Additional Material on TST Want to get kTST into Arrhenius form

k TST =
but so

k BT ' K h

K ' = eG

/ RT

RT ln K = G

k TST =

k BT G / RT kT = B e S / R e H / RT e h h

because G = H TS NOW:

H = E + nRT

Where n = # molecules in TS molecularity of reaction (molecularity: eg: unimolecular, bimolecular, etc) e.g.:

n = 1

-2

-1

So:

kT k TST = B e S / R e1 e E / RT h   

k TST =

BT m

e E

/ RT

5.62 Spring 2004 (where m=1)

Lecture #34, Page 11

Again, theory predicts a temperature dependence to the pre-exponential factor which is difficult to observe experimentally unless the rate constant is measured over a wide temperature range (At least a factor of 5). Now:
k d ln B e S / R e1 + ln T + e E / RT h = dT

d ln k TST dT E 1 = + T RT 2

Arrhenius :

d ln k Eact = dT RT 2

d ln k d ln k TST = dT dT Eact 1 E = + RT 2 T RT 2 Eact = RT + E

5.62 Spring 2004

Lecture #34, Page 12

Treat H2 as an atom a hard sphere of mass 2 (no rotation, no vibration) Treat H2F as a diatomic molecule (One unknown parameter, dH2-F) H2 + F H2F HF + H With these assumptions

kT k B h

E / RT (qtrans / N) H2 qrot e F ( q / N )( q / N ) trans trans

Note: no vibrational partition function for H2F is included because the one vibrational mode for the pseudo diatomic molecule transition state has become the reaction coordinate.

2 m + m k T 3/ 2 F) B ( H2 kT h3 N k B h 2 m k T 3/ 2 [ 2 m k T ]3/ 2 H2 B F B 3 hN h3 N

2 8 I k BT E / RT e 2 h

The reason there is no rotational or vibrational partition function for H2 is not that we are assuming the high-T limit, but rather H2 is treated as if it were an atom. Now
2 I = dH , = 2 F

mH 2 mF mH 2 + mF

8k T mH + mF k =N B 2 mH 2 mF

1/ 2

2 dH

2 F

e E

/ RT

5.62 Spring 2004

Lecture #34, Page 13

This looks identical to the collision theory result. Collision theory is not based on thermodynamics.

CT

8k T 2 = B d AB e E0 / RT

1/ 2

Calculate value for kTST in limit of collision theory assumptions (ie: whwat fraction of collisions are effective because they have sufficient translational energy along the line of centers?):

=1

2 dH

2 F

= 3 1019 m 2

k = 3.4 108 e E

/ RT

m3 / mol s

Compare to kTST

k TST = 3.9 107 e E

/ RT

m3 / mol s

kTST is smaller because it reflects the more restrictive collinear steric requirement. KCT is an upper bound because collision theory treats reactants as spheres with no favored direction of approach (but with explicit requirement of the effective collision energy).