CPMD: perspectives and industrial
applications
Mauro Boero
Institut de Physique et Chimie des Matériaux de Strasbourg, UMR
7504 CNRS-UDS, 23 rue du Loess, BP 43, F-67034 Strasbourg,
France and CREST, Japan Science and Technology Agency,
Kawaguchi, Saitama 332-0012, Japan, and JAIST, Hokuriku,
Ishikawa, Japan
Work done in collaboration with BASF A.G.
and Sumitomo Co.
Outline
 Main interests and computational details
 MgCl2 active surfaces
 TiCl4 and Ti2Cl6 catalytic species
 Active centers and complex formation: stability
problems under real reaction conditions
 a-Olefins polymerization: ethylene and propylene
 The role of a donor phthalate (few notes)
 Conclusions and perspectives (?)
Main interests
• Ziegler-Natta (ZN) catalysis is by far the most
important industrial process in the production of
polyolefins with high degree of stereoselectivity
• The reaction occurs at room temperature with a
very high reaction rate and low amount of catalyst
• Experimental probes fail in recovering the
microscopic picture due to the very fast reaction
and the low percentage of active sites
• Quantum dynamical simulations can be a viable
tool to study in an unbiased way and on affordable
time scale active sites and the reaction pathway
• Annual Worldwide Production (2008)1 : 45 million tons
• Share of consumption by region (2007)2 :
China — 23%
Western Europe — 18%
Middle East/Africa — 9%
Central/South America — 5%
North America — 18%
Asia/Pacific — 16%
Japan — 11%
• Key Products: Packaging, textiles, fibers, automotive components, cups,
cutlery, housewares, appliances, electronic components, carpeting, photo
and graphic arts products.
• Top 10 world producers by 2008 market share (in descending
order)3 : LyondellBasell, Sinopec Group, Saudi Basic Industries Corp.
(SABIC), PetroChina Group, Reliance Industries, ExxonMobil, Borealis,
Total PC, Ineos, Formosa Plastics
1 Source: ChemSystems
2 Source: Townsend Solutions
3 Source: Chemical Market Associates, Inc. (CMAI)
Polyolefins items produced routinely in
industries and laboratories
MgCl2 simulation cells:
• Surface (100), orthorombic, 32 formula units,
17.673 x 14.560 x 28.000 Ǻ 3
• Surface (110), monoclinic, 30 formula units,
19.095 x 12.522 x 28.000 Ǻ 3
a b = 70.2o
• Surface (104), orthorombic, 48 formula units,
14.560 x 21.711 x 28.000 Ǻ 3
MgCl2 bulk crystal structure
Space group: R3m, a = 3.640 Ǻ, b = 17.673 Ǻ
MgCl2 (110) surface
• Unrelaxed:
Mg-Cl = 2.526 Ǻ
ClMgCl = 180o
• Relaxed:
Mg-Cl = 2.322 Ǻ
2.403 Ǻ
ClMgCl = 154o
DE = 3.4 kcal/mol Mg
Deposition of the TiCl4 catalyst
Mononuclear Ti centers
Octahedral Ti 6-fold on
MgCl2 (110) surface as
proposed by Corradini
and co-workers.
Ebind= 40.3 kcal/mol
5-fold Ti site on MgCl2
(110) surface obtained
from CPMD
Ebind= 29.4 kcal/mol
[JACS 120, 2746 (1998)]
Active mononuclear centers
Active Corradini center:
Active 5-fold center:
removal of 1dangling Cl
substitution of a Cl with
+ substitution of a Cl with a methyl group.
a methyl group.
Free energy calculation
• Select the reaction coordinate x to be monitored.
Our choice: x=|C1-Ca| (olefin-chain distance)
• Add to the Car-Parrinello lagrangean LCP the
constraint x: LCP a LCP+ lx(x-x0)
• Compute the ensemble average < lx > according
to the Blue Moon prescription
• Integrate the constraint force along the sampled
path
see M. Sprik, G. Ciccotti, J. Chem. Phys. 109, 7737 (1998)
First insertion of ethylene
Main phases of the insertion
 The p-complex (left), the transition state (center)
and the final product (right)
 Reaction coordinate: x = |C1-Ca|
 The reaction is a-agostic assisted
Ethylene polymerization from mononuclear Ti
• ELF of the main steps of the ethylene insertion: the pcomplex (right), the transition state (center) and the final
product (right). ELF=0: blue, ELF=1: red
• ELF is projected on the plane containing the C1 and C2
carbon atoms (grey) of the ethylene and Ti (purple).
Catalysis of polyethylene: energetics
Corradini
6-fold site
5-fold
site
Experiment:
Barrier = 6-12 kcal/mol
Product = -22 kcal/mol
Second insertion of ethylene
Second insertion of ethylene
 p-complex (left), transition state (center) and final
product (right) in a single snapshot
 Reaction is b-agostic assisted
Isotacticity: regular polymer where each unit
has the same orientation
Propylene
molecule
Isotactic (stereoregular)
polypropylene
Q.: How can the Ziegler-Natta Ti catalyst produce isotactic
polypropylene ?
Possible olefin orientations for propagation
4 possible insertions, but only (a)
can give a barrierless complex
and a transition state lower by
~7-8 kcal/ mol with respect to
(b), (c) and (d).
This propylene enantioface has
the minimum steric hindrance
Propene polymerization from mononuclear Ti
The computed propene 1,2 insertion is
Isotactic. The barrier is 10.5 kcal/mol
(exp.9.5-12) with a final gain of 16.7
kcal/mol below the p-complex.
Cldangling-Mg = 3.03-3.22 Ǻ
Complexation and insertion energies for
ethylene and propene
Ethylene
Propene
(1,2)
Propene
(2,1)
Insertion
Product
p-complex
(kcal/mol) (kcal/mol) (kcal/mol)
-8.4a,-6.5b +12.7a,+6.7b -5.8a,-23.3b
-3.6b
+10.5b
-16.7b
+5.6b
+16.2b
(-1.0)b
a = Corradini site, b = 5-fold site
Donor di-n-butyl phthalate
• Ability to coordinate to the support also in presence of Ti.
O1-O2 = 2.7-3.9 Ǻ (phthalates, diethers)
• Absence of secondary reactions with the catalyst
• Absence of secondary reactions with the Metal-C bond and
the growing polymer
Interaction of a donor with the support
(100)
Mg1-Mg2
(Ǻ)
3.674
Mg1-O1
(Ǻ)
2.150
Mg2-O2
(Ǻ)
2.246
Ebind
(kcal/mol)
21.9
(110)
6.360
> 4.0
2.085
8.5
Interaction of the donor with the active
Corradini center
• Ti-Mg1= 3.781 Ǻ, close to
Mg-Mg distance on (100)
• Mg1-O1= 2.203 Ǻ
Ti-O2 = 1.942 Ǻ
• Binding energy = 20.1
kcal/mol
• The center is poisoned by
the donor
• No interaction with 5-fold
Conclusions (and perspectives)
• The role and the relative importance of various
MgCl2 active surfaces has been investigated
• Interaction of Ti species with the support has been
studied
• The polymerization reaction has been investigated
and the reaction pathway elucidated
• The problem of the stability of the active sites in a
realistic system has been inspected
• The possible role of a donor phthalate has been
addressed at a dynamical first principles level
Acknowledgements
•
•
•
•
•
•
•
Hors Weiss, BASF AG
Stephan Hueffer, BASF AG
John Lynch, BASF-TARGOR
Akinobu Shiga, Sumitomo Co.
Shinichiro Nakamura, Mitsubishi Co.
Minoru Terano, JAIST
Hans-Joachim Freund, Fritz-Haber-Institut MPG
Related publications:
J. Am. Chem. Soc. 120, 2746 (1998)
Surf. Sci. 438, 1 (1999)
J. Am. Chem. Soc. 122, 501 (2000)
J. Phys. Chem. A 105, 5096 (2001)
Macromol. Symposia 173, 137-147 (2001)
Mol. Phys. 100, 2935 (2002)