Note: Descriptions are shown in the official language in which they were submitted.
1 It is kno~n in the ~rt that alpha-olefins can be
2 polymeriz~d in the presence of cataly~ic systems containing
3 solid TiC13 or solid TiC13 m~xed with o~her solid metallic
4 halides7 usually as cocrystalLine eutectics activated by an
organic compound of aluminum.
6 These solid compositions containing titanium trl-
7 chloride can be prepared by a number of different processes.
8 One of these processes is the reduction of TiG14 by hydrogenO
9 Another process consists of reducing TiC14 with a metal such
as aluminum where the cocrystalline material would be alum-
11 inum chloride. Another process preferred for several reasons
12 is reduction of TiC14 with alkyl aluminum (halides)0 The
13 TiC13 product which then is normally in ~he brown be~ form
14 contains either aluminum chloride or alkyl aluminum chloride
by-products, or both~ associa~ed with the TiC130 For op-
16 timum catalytic effects, it is preferred that this brown
17 material be converted to the purple form by either heating
18 or using excess titanium tetrachloride.
19 When alpha-olefins, for example propylene, are
polymerized with these ca~alysts, commercially undesirable
21 amounts of amorphous polypropylene are formed along with the
22 desirable isotactic crystalline pol~propylene. It is well
23 known in the ar~ that third components can be added as com-
24 plexing agents to titanium trichloride catalysts to improve
the isotacticity of the added crystalline polyolefins, al
26 though usually at the eost of reduced efficiency of the
27 polymerization reaction.
28 Thus, Boor & Jordan have descr~bed how titanium
29 trichloride can be imprvved by -the addition of Lewis bases
(J.Boor, Jr., "Active Site in Ziegler Catalysts", page 115
31 in "Macromolecular Reviews, Vol. 2"; see also D 0. Jordan,
32 "Ziegler Natta Polymerization" in "The Stereochemistry oE
- 2 -
~ ~ 6~
1 Macromolecules, Vol. l", edited by A~ Do Ketleyg 1967,
2 Marcel Decker, Inc.).
3 Grignard reagents comprising magnesium compounds
4 complexed with ethers have been used to reduce TiCl4 to
TiCl3 for ethylene polymerizations~ See UOS~ 3~801~558O
6 But the ethers are only used to solubili2e the Grignard
7 reagent which is otherwise hydrocarbon-insoluble.
8 That TiCl3 catalysts can be improved by the addi-
9 tion of Lewis bases, for example ethers, and/or TiC14 treat
ments has been disclosed in German patent DT-2213086~
11 It is demonstrated in that patent how TiCl4 can be reduced
12 with diethyl aluminum chloride to yield a brown reduction
3 product containing titanium trichloride. This brown reduc
4 tion product separated from the reaction medium was subse-
quently treated with a special ether, and subsequent to
16 the ether treatment the reaction product was separated and
17 then treated with excess titanium tetrachloride to form a
18 purple catalyst, which when separated from the resction medi-
19 um could be activated with diethyl aluminum chloride to
yield a catalyst active for the stereospecific polymerization
21 of propylene.
22 Thus, while it is well known that Lewis bases, and
23 specifically ethers, can be used to treat preformed catalysts
24 to improve the stereospecificity and the actîvity of said
catalysts, actually carrying out the reduction step for re-
26 ducing TiC14 to TiCl3 in the presence of a Lewis base has
27 not been reported nor is it obviousO
28 A highly efficient process to make a catalyst of
29 outstanding quality results from the reduction of TiC14 to
TiCl3 in the presence of Lewis bases which complex with
31 the reducing agents.
32 It has been found~ according to ~his inven~ion,
~3.~ 6 ~ ~ ~
1 that titanium tetrachloride can be reduced with alkyl alum-
2 inum (halides) in the presence of Lewis bases to produce
3 highly stereospecific active catalysts.
4 The advantages of the process of the invention
are~ (1) the activity of the reducing agent is tempered,
6 thereby allowing greater control of the reduction s~ep 1,
7 (2) the entire process sequence is simplified since the
8 entire reduction and catalyst preparation is performed in
9 only one reactor by simple sequential addition of reagents;
o and (3) by this process highly stereospecific highly acti~e
11 catalysts having properties far superior to those of commer-
12 cially available catalysts are formed~
13 The catalysts are preferably fo~med in a novel
14 two-step process, whereby in the first step TiC14 is added
to an ether/aluminum alkyl (halide) complex or vice ~ersa
16 at a temperature low enough to control the reduction step,
17 and after the reagents have been mixed the temperature is
18 increased to effect complete reduction maintaining a speci-
19 fied stirring speed so as to obtain the catalyst particles
in a controlled form. Subsequent to the completed reduction
21 the second phase of the catalyst preparation is enacted by
22 adding TiC14 to the reaction mixture~ and then heating the
23 catalyst and TiC14 at a specified temperature for a time
24 long enough to conver~ the catalyst to the purple form.
Thus, catalysts are prepared by complexing alum-
26 inum alkyl halides with the general formula AlRnX3_n where
27 R is a hydrocarbon radical containing from 1 to 8 carbon
28 atoms, preferably 1 to 4 carbon atoms (where the best re-
29 sults are obtained when R is selected from ~he group of al-
kyl, aryl radicals) and X is a halogen selected from chlor-
31 ine, fluorine, bromine, and iodine where the best results
32 are obtained when X is chlorine and n is any number between
0 and 3. Preferably, n is 1.5 to 2.5, with the best results
obtained when n is equal to about 1.7. It i5 also understood
that the reducing agent can be of the class of aluminum alkyl
alkoxides, where in the above general formula X would be OR
instead of halogen; or that the reducing agent is of the class
of the organo aluminum polymer compounds whereby two or more
of the above-described aluminum alkyl reducing agents are
joined by oxygen, nitrogen, sulfur, or methylene bridges.
The Lewis base complexing agent that is complexed to
the aluminum alkyl reducing agent can be any compound known
in the art which complexes with Lewis acids. Thus those
compounds containing one or more atoms or groups having one
or more pairs of free electrons capable of effecting coordina-
tion with titanium and aluminum alkyl (halides) are generally
usable. Specifically among the atoms capable of donating one
or more pairs of electrons are those atoms of Groups V and VI
of the Periodic Table for example oxygen, sulfur, nitrogen,
phosphorous. As representative examples of compounds contain-
ing groups capable of furnishing one or more pairs of electrons,
mention rnay be made of ethers, thioethers, phosphines and
amines.
It is preferred to use as complexing ayents those
compounds having the general formula ROR', RSR', R(R')~R"
where R, R' and R" are hydrocarbon radicals containing from
1 to 15 carbon atoms. The best results are obtained when
R, R' and R" are branched hydrocarbon radicals containing from
2 to 8, preferably from 4 to 6, carbon atoms.
The complexing agent is mixed with the reducing agent
in the ratio of a molar excess up to 5 to 1 preferably in the
ratio of an excess up to 2 to l. And it can be premixed or
can be mixed in the reactor and generated ln situ.
It is well k~own that such a complexed reagent has different
reducing powers and behaves differently from the pure reducing
agent by itself. The TiC14 may be added to the complexing agent
or vice versa the TiC14 can be added to the complexed aluminum
alkyl reducing agent. The several reagents are used in the
following ratios expressed by the general formula:
complexing agent(X): aluminum alkyl reducing agent(y):
TiC14(z) where x, y and z can have values from 71 to 5, 1-5,
1-20, respectively, or preferably ~1 to 2, 1, 1-2, as long as
x:y is in molar excess and X:y+z is less than 1, or most
preferably 1.5, 1, 1, respectively.
While the reduction can be effected at temperatures
from -80 to +50C., it is preferred to reduce TiC14 at tempera-
ture from -30 to ~15C., or most preferably from -10 to ~5C.
While many solvents can be used for the reduction, it
is most preferred that the solvent is a nonreactive solvent
such as the paraffinic or alkyl aromatic solvents. Thus, the
type of inert solvents that are preferred are those selected
~ from the paraffinic hydrocarbons containing from 5 to 12 carbon
atoms. Most preferred solvents are taken from the group of
pentane, hexane, heptane or isooctane.
The concentration of reagents and the inert diluent
can vary from 0.5 to 4 molar but best results are obtained
when the concentration is between 2 and 3 molar.
A wide variety of stirring techniques can be effected
during reduction leading in all cases to good catalysts. For
better control of the final catalyst product, it is preferred
to stir the reaction mixture in a smooth manner during the
reduction and warm-up at a rate sufficient to yield catalyst
particles having all approximately the same size. Pre~erred
stirring rates are from 50 to 600 rpm;
-- 6 --
l best results are obtained when the stirring rates are be-
2 tween 150 and 400 rpm.
3 The final reaction mixture is warmed up from the
4 reducing temperature to an aging temperature which can be
varied from 25 to 90C., with best results obtained when
6 the aging temperature is from 50 to 65Co The rate of warm-
7 up can vary from Ool to 3Co per minu~e where the preferred
8 results are obtained when the warm-up rate i5 from 0~5 to
9 1C. per minuteO
The catalyst is held at the aging temperature from
ll 1/2 to 6 hours preferably from 105 to 205 hours~
12 Subsequent to the aging, TiC14 is added to the re-
13 action mixture in an amount where the TiC14 to TiC13 mole
14 ratio can vary from 005 to 20 or preferably from 1 to 10 or
most preferably from 1 to 3-5 The TiC14 can be added in an
16 undiluted form which simplifies the process and the concen-
17 trations of reactants and presence of diluent can be varied
18 so as to vary the concentration of the TiC14 in the reaction
19 mixture from 5 to 75 vol. %, preferably from 25 to 50 v310%.
The catalyst after such treatment with TiC14 is
21 heat soak treated from 1/2 to 10 hours, preferably from 1
22 to 2 hours at temperatures which range from ~25 to 90C0,
23 preferably from 50 to 65Co
24 The thus prepared catalyst is separated from the
reaction mixture and washed by decantation or filtration
26 with the diluent used in the preparation or with other un-
27 reactive hydrocarbonsO
28 It is understood that for simplification also ex-
29 cess TiC14 can be trea~ed with the aluminum alkyl ether com-
plex so that the subsequent TiC14 treatment described above
3l does not need to be effected
32 It is also understood that the titanium tetra-
~ 7.~
chloride can be premixed with othex transition metal ha]ides
of Groups IIIB~ IVB, VB before reduction and separation to
thereby generate a solid reduction product containing mixtures
of transition ~etal halides, where these mi~tures will generate
polyolefins having different properties. Finally, it is also
understood that the TiC14 reduced as described above can be
reduced with and by metal alkyl (halides) other than aluminum
alkyl halides. Thus metal alkyl halides of metals of Groups IA,
IIA, IIB, IIIA and IVA can be used.
It is known that both ether and TiCl4 must be present
in the catalyst to effect activation, and since enough ether
must be present to complex at least all of the aluminum compound
in the catalyst, it would be preferable to have a slight excess
of ether in the reduction mixture; however, the ether should
not exceed the molar quantities of aluminum alkyl compound and
TiCl4, for then significant reduction would not be expected.
Accordingly, within critical narrow ranges of ether: aluminum
alkyl compound TiCl4 mole ratios, highly active, stereo-specific
catalysts were obtained.
If in the reduction step, using Et2AlCl, the ratio of
ether:AlEt2Cl did not exceed l or if the ether:LTiCl4 -~ AlEt2Cl~
equals l or more, conversion of the brown TiCI3 to the purple
form by excess TiCl4 is not possible. See the table following
for a summary of the results using different ether ratios
during reduction.
Catalyst Color
Reducing Agent Ticl4:AlEt2cl:r)IpE After TiCl~Treatment
., . . . .. . __ __ .
AlEt2Cl.2.0 DIPE l:0.5 :l Purple
AlEt2Cl.l.3 DIPE l:0.75 :l Purple
AlEt2Cl.2.0 DIPE l:l :2 Brown
30 AlEt2Cl.l.0 DIPE l:0.5 : 0.5 Brown
AlEt2Cl.l.0 DIPE l:l :l Brown
-- 8
~.
~ 6 ~S~.~
l Since the first step in the reduction yields
2 AlEtC12, and since an AlEtcl2~Base complex might be con
3 sidered too weak to reduce TiC14, one might expect the
4 yield of TiC13 to be limited to 50% of theoretical. See
the following equations. TiC14+005 AlEt2Cl R2Q - O.S
6 TiC13~OD5 AlEtC12-R20 + 005 TiC14 0~5 TiC14+0 5 AlEtC12~R20 -
7 005 TiC13+0-5 AlC13 R200
8 This concern is reasonable, for while AlEtC12
9 reduces TiC14, the AlEtC12R20 complex with TiC14 yields
only a clear green solutionO However, the yield of solid
ll catalyst actually obtained indicates that both ethyl groups
12 in AlEt2CloR20 function to reduce TiC13c Possibly TiC13
catalyzes the reduction of the AlEtC12oR~O~TiC14 complex
in the same way that TiC13 catalyzes the reduction of TiC13R.
Another possibility is that TiC140R20 complex reduces faster
16 than TiC14, and that therefore in cases where the former
17 can be formed, reduction proceedsO
18 It has been observed that the ethero AlEt2Cl ratio
19 must exceed 1 for beta ~o delta TiC13 co~version to occur,
but the reason for this critical ratio is not understoodO
21 At ratios below 1, free aluminum alkyl would be present,
22 but why that should deposit TiC13 in a form less susceptible
23 to rearrangement is difficult to explain, especially since
24 TiC13 catalysts prepared without ether can be converted to
the purple delta form by excess TiC140 See the following
26 equations:
1) Heat Brown beta
27 TiC14 + 0.75 AlEt2CloO~5 R20 2~ XS TiC14 TiC13
28 TiC14 ~ 0075 AlEt2cl 2~ XS TiC14 T~C~2
B 29 One reaction that occurs at low R200AlEt2Cl ratios
is the disassociation of the AlEtC12 by-productl yielding
31 the AlC13R2o complex.
32 Triethyl aluminum can also in a similar manner be
_ g _
l~ S9
complexed with a Lewis base thereby decreasing the reducing
2 power of the more active reducing agentO This lower activ-
3 ity is only observed when excess ether is present complexing
4 all the triethyl aluminum.
The lower reduction rate observed with complexed
6 Et2AlCl suggests a process advantage one might expect using
7 AlEt2Cl-R20 rather-than- AlEt2Clo In large scale catalyst
8 preparation, temperature control becomes difficulto The
g reduction of TiC14 with complexed aluminum alkyl should be
lo gentler and therefore easier to controlO
11 Dimethyl aluminum chloride and trimethyl alumin-
12 um reduce TiC14 at a lower rate than the ethyl analogs;
3 however, complexed with ether, days are required for re-
14 duction to be effectedO
Lewis bases have been described in many patents
16 and publications, such as U.S0 3,116,274, UOS0 3~825,524
17 and on page 31 of a book entitled SEMIMICRO QUALITATIVE
18 ANALYSIS, by William CO Oelke, publi4hed in 1950 by D.C.
19 Heath & Co.
In addition to the reduction step, the complexes
21 of the invention can also be used as a cocatalyst for either
22 the TiC13 catalysts made by the process of the invention
23 or for TiC13 and other Ziegler transition metal catalysts.
24 The invention is further illustrated by the follow-
ing examples:
26 EXA~LE 1
27 To a 250 ml flask is added 24~5 cc (54.6 mm) of
28 a 2.23 molar diethyl aluminum chloride solution in iso
29 octanec Then, 17.8 cc (27.3 mm) of a 1053 molar solution of
ethyl aluminum dichloride in hexane is added. The reducing
31 agent mixture is cooled to 0 and 22~4 cc (10902 mm) of di-
32 isopentyl ether is added dropwise. 6 cc (5406 mm) of 100%
- 10 ~
l TiC14 is added at a rate of 0O15 cc per minuteO At the
2 end of the titanium tetrachloride addition, the reaction
3 mixture is held for 60 minutes and then warmed to 65C. at
4 a rate of 1 per minute~
The reaction mixture is held for 1 hour at 65Co,
6 cooled to 35C., and 60 cc (546 mm) of 100% TiC14 is added
7 at a rate of 2 cc per minuteO The stirring rate is cut
8 back from the 300 rpm used during the reduction to 100 rpm
9 and the solution is warmed to 65C., held there for 2 hoursO
The catalyst examined on the microscope has turned purple
ll and has a narrow particle size distribution where the aver
12 age particle diameter is 12 micronsO The catalyst is sepa-
3 rated from the reaction medium by filtration and washed
4 three times with 100% heptaneO
To a 1,000 cc flask containing 500 cc of normal
16 heptane at 65Co is added 4~37 cc of a 1~6 molar solution
17 of diethyl aluminum chloride in normal heptaneO Then 0070
18 grams of the catalyst as prepared above is added to the re-
19 action mixture which has been saturated with propyleneO
For 2 hours propylene is passed through the
21 stirred solution and polypropylene is generated by the cata-
22 lystO At the end of 2 hours addition of propylene is re-
23 placed by addition of nitrogen and the polymer slurry is
24 stirred for 12 hours with 1000 cc of isopropyl alcohol and
then filteredO The polymer cake is washed with 100 cc of
26 alcohol and the washings added to the filtrateO The fil
27 trate is evaporated to yield 5 o 94 grams of waxy polymer and
28 catalyst residueO
29 The polymer is dried at 65~Co and weighed to
yield 87.36 grams of polymer of which 98099% is insoluble
31 in boiling heptaneO After correction for the catalyst
32 residue in the filtrate, the catalyst efficiency is calcu-
- 11 -
~ 5 ~
1 lated to be 125.4 grams of polymer per gram of catalyst for
2 a 2 hour polymerization, 0.49% of the polymer is called
3 waxy polymer, that is, polymer soluble in the heptane-iso-
4 propyl alcohol slurry and 98.5% of the polymer is insoluble
in boiling heptaneO
6 The polymer particles have a narrow particle size
7 distribution, flow well and have an average diameter of 75
8 micronso
9 By comparison, commercial TiC13 M obtained from
Stauffer Chemical would have a catalyst efficiency of 42
11 of which 92~5% of the polymer would be insoluble in boiling
12 heptane. Furthermore, the polymer would have a w;de par-
13 ticle size distribution.
4 By further comparison, a catalyst prepared by re-
duction of TiC14 with diethyl aluminum chloride, and heated
16 to convert the catalyst from the brown to the purple form
17 would have a polymerization catalyst efficiency of 38 and of
18 which ~.5% of the polymer would be insoluble in boiling hep-
19 taneO
EXAMPLES 2-4
21 Other catalysts were prepared according to the
22 data illustrated in Table 1 below for Examples 2-4 where
23 the concentration of TiC14 and the etheroaluminum alkyl
24 ratio was varied~
EXAMPLES 5-11
26 Catalysts were prepared by the reduction of TiC14
27 with diethyl aluminum chloride complexed with diisopentyl
28 ether or ethyl ether and as shown in Table 2 below, improved
29 catalysts having high efficiencies and high percent heptane
insolubles were obtainedO
31 EXAMPLES 12_14
32 TiC14 was reduced with triethyl aluminum complexed
- 12 -
with diisopentyl ether and as ~ tstrated in Table 3
2 below yielded catalysts that had high ca~alys ~ e~ic
3 iency and high percent heptane insolubles~
~ ~3 -
o ~i
.
. ~ I . .
~ -
v ~
1~1 ;~ ~I m ~
~ ~ . I
~l ~ ~
~ ~ ' Lt~' ' '' ' '
!~
e ~ ~!
V . . ~ ~ ,
~ ~
~ ~ uo u.~
~1 ~ Q
~ ~3
~ ~ n ~ a
. ~ u~
~ _ ~ h ~ E~ 3 ~ ~3 i~
~ ~ s~ l
dP
;~
~` J~
~ o ~
~ o ~i o
~ . .. I
.~,q . . ..
. ~ ~
~ ~ o~ o
. t . .
.
.. .
~ ' ~ ~
o , ~ ~ ~
h~ k;~ ~
Q ~ ~ ~
Y~-~ . . .
. ~ ,.~ ,, ,
:~ .~ ~ , . ~.
Q
~ . ~ - . i
'k
~1 o o o o' o 'o o
. ~; .. .. .. ~. .. .. .. .
': . ~
.~ ;~
.~ .~ .
.~ I !~
I ~ ~ , ~ ~ ~ ~ W
.1 N rl _I ~1 ~1 _i
~ ~l iv iv iv ~ iv
z ~ ~
1) ~ 0 ~
; y ~
o
~3 ~ P E~ ~ a ,~ P a a a, ~ ~ ~ a
:~ p, ~ rc J ~ ~ 3
5~ ~
~ --15-- ~ ~
¦ ;~a~ 161.i~3
~r ~ , . . .
~¦ ~,a ¦ 1- 0
~ . ,.
N W
. ~ ,,
.
~ ,. ' ' ~ _
~ 1 ~0~
,. .
. ~ ~ ' g
~o~ ~E~ ,4E~
o~ o~ ~ o~ ~
- z ~ z ~ ~ z ~
--16--
. .-- . . .
1 Although diisopentyl ether is an excellent performer
2 and had been used in many experiments of the invention,
3 it is hereby disclosed that another excellent performing
4 ether is benzyl ether.
It is to be emphasized-that the ether/ethylene
6 aluminum chloride ratio must exceed 1 in order for the
7- TiC14 to be converted to the purple TiC13 after the
8 final TiC14 after-treatment.
9 It is also apparent that the ether aluminum ethyl
chloride complex actually;changes the mechanism of
11 reduction and`the properties of the final catalyst
12 product.
13 Another important observation regarding the complex
14 of the invention is that the various aluminum alkyl or
aluminum alkyl chloride complexes are quite different
16 depending on the Lewis acid strength of the aluminum
17 alkyl. Thus, the strongest reducing agents go in the
18 order of aluminum triethyl, followed by aluminum diethyl
19 chloride, followed by ethyl aluminum dlchloride. But
the Lewis acidity strength of these reducing agents proceeds
21 from strong to weak, e.g., from ethyl aluminum dichloride,
22 which is stronger than diethyl aluminum chloride, which
23 , is stronger than triethyl aluminum.
24 The ethyl aluminum dichloride complex with diisopentyl
ether is the least effective reducing agent.
26 The best catalyst results are obtained when using a
27 mixture of ethyl aluminum dichloride and diethyl aluminum
28 dichloride with diisopentyl ether.
29 When diethyl aluminum chloride is used by itself,
as pointed out above, the ether diethyl aluminum chloride
31 ratio must exceed 1 in order to obtain the desired
32 purple delta TiC13.
-17-
s9
1 Tri~.thyl alllminum complexed ~ith dii.sopentyl ether
2 was observed to be the most active reducing agent yielding .
3 very high percentage yields at temperatures around 0C.,
4 but suffering from a very broad particle size distr1bution.
.
-' -
.' ,: ' . ' . I
- ' ' ~'
.
., . , ~
. . I
.
. .
, . , ' ' ~ I
,
.
.
, '
-18-
