- Email: [email protected]
Propylene polymerization behavior of Ti-containing mesoporous silicas
Studies in Surface Science and Catalysis, volume 158 J. (~ejka,N. Zilkov~iand P. Nachtigall (Editors) 9 2005 ElsevierB.V. All rights reserved.
1437
Propylene polymerization behavior of Ti-containing mesoporous silicas Y. O u m i a, S. T a k a s h i m a a, A. Hanai a, H. N a k a j i m a b, K. Y a m a d a b, S. Hosoda b and T. Sano a
aSchool of Materials Science, Japan Advanced Institute of Science and Technology, Tatsunokuchi, Ishikawa 923-1292, Japan ; E-mail: [email protected] bpetrochemicals Research Laboratory, Sumitomo Chemical Co. Ltd., Sodegaura, Chiba 299-0295, Japan Ti-containing mesoporous silicas (Ti-MCM-41, Ti-MCM-48 and Ti-SBA-15) were prepared by the post-synthesis with Ti(OC4H9)4 and their propylene polymerization behavior was investigated. It was found that these Ti-containing mesoporous silicas combined with Al(i-C4H9)3 provide the isotactic polypropylene. Polypropylenes outside and inside the mesopores of Ti-containing mesoporous silicas had different characteristics despite being polymerized concurrently. The crystallization of polypropylenes confined in the mesopores was considerably depressed due to the limited space. 1. INTRODUCTION Recently, there has been a great deal of interest in the behavior of polymers under confined geometries for understanding and controlling the properties of polymers on micrometer and nanometer scales. Several inorganic porous materials such as layered silicate, mesoporous silica and porous glass are used as a nano-reactor. However, there are only a few papers concerning polymerization of olefins such as ethylene and propylene with mesoporous materials, in which organometallic complexes are impregnated or grafted on the surfaces of mesopores [ 1-3]. Since the discover of ordered mesoporous silicas such as MCM-41, MCM-48 and SBA-15, much effort has been paid to incorporation of various metals into these framework structures by the direct hydrothermal synthesis and the post-synthesis methods for heterogeneous catalysis and adsorption. It is well known that Ti- and V-containing mesoporous silicas are active for selective oxidation of a wide variety of organic substrates with the environmentally friendly oxidant [4-6]. From such viewpoints, we have studied the olefin polymerization in mesopores, and very recently found metal-containing MCM-41, especially Ti-containing MCM-41 combined with alkylaluminums can provide isotactic polypropylene with broad molecular weight distribution [7,8]. However, the detailed olefin polymerization behavior in mesopores is still an open question. In this paper, therefore, several Ti-containing mesoporous silicas with different mesoporous structures such as Ti-MCM-41, Ti-MCM-48 and Ti-SBA-15 were prepared and used to polymerize propylene. By analyzing the characteristics of polypropylenes both outside and inside the mesopores, the influences of the mesopore structure of Ti-containing mesoporous silica on the polymerization behavior was investigated in detail.
1438 2. EXPERIMENTAL
2.1. Preparation and characterization of Ti-containing mesoporous silicas Ti-containing mesoporous silicas were prepared by the post-synthesis method. The parent siliceous MCM-41, MCM-48 and SBA-15 were prepared following the procedures described in the literature [9-11 ]. The siliceous mesoporous silicas were calcined at 500~ for 10 h to decompose the surfactant. 1 g of the mesoporous silica calcined at 500~ for 10 h was dispersed in 10 ml of dry toluene containing 3 mmol of Ti(OC4H9)4. The mixture was refluxed at 115~ for 24 h. The product was filtered, washed with dry toluene several times, dried at room temperature and then calcined at 500~ in air for 5 h. Identification of the products was carried out by X-ray diffraction (XRD, Rigaku RINT-2000). Elemental composition was determined by X-ray fluorescence (XRF, Philips PW-2400). Textural properties (BET surface area, pore diameter, pore volume) were evaluated by nitrogen adsorption at -196~ (Bel Japan, Belsorp 28SA). UV-Vis spectra were recorded on a Hitachi U-3310 spectrometer. 2.2. Polymerization of propylene Propylene polymerization was conducted at 40~ in a 100 cm 3 stainless steel autoclave equipped with a magnetic stirrer. After the reactor was filled with nitrogen, measured amounts of Ti-containing mesoporous silica evacuated at 400~ for 8 h, toluene and Al(i-C4H9)3 were added to the reactor in this order and the mixture was kept standing at room temperature for 15 min. The reactor was evacuated at liquid nitrogen temperature, and the 7 dm 3 of propylene was introduced. Polymerization was started by quickly heating the reactor up to the polymerization temperature (40~ The polymerization reaction was terminated by adding acidified methanol (5 wt% HC1). For characterization of the polypropylene outside mesopores of Ti-containing mesoporous silica (ppout), the as-synthesized polypropylene was extracted with the boiling o-dichlorobenzene (ODCB). Also, to obtain the polypropylene inside mesopores (pp,n), the
t
Extraction with ~ bc ling ODCB Y
(
(d)
Alkali treatment
Extraction wit boiling ODCB Fig. 1. Schematic diagram of various treatments. (a) as-polymerized PP, (b) PP/Ti-containing mesoporous silica nanocomposite, (c) PP outside mesopores of Ti-containing mesoporous silica and (d) PP inside mesopores ofTi-containing mesoporous silica.
1439 PP/Ti-containing mesoporous silica nanocomposite was treated with a 12M NaOH aqueous solution at 120~ for 48 h and this was followed by extraction with ODCB (Fig. 1). The weight-average molecular weight (Mw) and molar mass distribution (Mw/Mn, Mn: number-average molecular weight) of the polymers were measured at 145~ by gel-permeation chromatography (GPC, Senshu Scientific SSC7100) using ODCB as a solvent. The melting points (Tm) of the polymers were measured on a Seiko DSC-220C calorimeter with a heating rate of 10~ min -1 in the temperature range of-40 - 200~ ~3C NMR spectra of the polymers were measured in 1,2,4-trichloro benzene/benzene-d6 (9/1 v/v) at 140~ using a Varian GEM-300 spectrometer operating at 75.4 MHz. Solid state 13C CP/MAS NMR spectra of the PP/Ti-containing mesoporous silicas were recorded using a zirconia rotor with a 7 mm diameter on a Varian VXP-400 spectrometer at 100.7 MHz. 3. RESULTS AND DISCUSSION 3.1. Preparation
of Ti-containing
mesoporous
silicas
The parent siliceous MCM-41, MCM-48 and SBA-15 exhibited the typical XRD patterns. MCM-41 and SBA- 15 exhibited three diffraction peaks ((100), (110), (200)), indicating the hexagonal framework structure. On the other hand, MCM-48 exhibited the diffraction peaks corresponding to the highly branched and interwoven three-dimensional channel system. Fig. 2 shows the XRD patterns of Ti-MCM-41, Ti-MCM-48 and Ti-SBA-15 prepared by the post-synthesis method. All samples gave slightly lower quality XRD patterns than those of the corresponding parents. The XRD patterns were found to be nearly free from crystalline TiO2.
(•i
(lOO) lo)2 oo)
(c)
.(211)
r~
~0~420)~(322)
~D
(b)
100)
0) i
0
I
_
(a)
4 6 2 theta (degree) Fig. 2. XRD patterns of (a) Ti-MCM-41, (b) Ti-MCM-48 and (c) Ti-SBA-15.
I
!
I
2
200
,
I
,
I
,
I
,
I
,
300
400 500 600 700 800 Wavelength (nm) Fig. 3. UV-Vis spectrum of Ti-MCM-48.
Table 1 Characteristics of various Ti-containing mesoporous silicas. Sample (Si/Ti ratio) BET surface area (m 2 g-l) Pore volume (cm 3 g-l) Pore diameter (nm) Ti-MCM-41 (22.4) Ti-MCM-48 (21.7) Ti-SBA-15 (22.5)
908 1,215 647
0.72 1.10 0.72
2.74 2.52 5.80
1440 This was also confirmed from UV-Vis analysis. As can be seen in Fig. 3, the UV-Vis spectrum showed an absorption band at ca. 220 nm due to a charge transfer between framework oxygen and tetrahedral Ti(IV) [12]. However, the spectrum also showed a shoulder around 330 nm corresponding to octahedrally coordinated Ti species, indicating the presence of trace amount of octahedral TiO2. Characteristics of these Ti-containing mesoporous silicas are listed in Table 1. 3.2. Polymerization of propylene Propylene polymerization was carried out using the Ti-containing mesoporous silicas prepared. All of the Ti-containing mesoporous silicas combined with A1 (i-C4H9)3 displayed activity and gave selectively isotactic polypropylene (PP). The order of the polymerization activity was as follows: Ti-MCM-48>Ti-SBA-15>Ti-MCM-41 (Table 2). This suggests that the pore size and structure of mesoporous silica affect the polymerization activity; three-dimensional structure of Ti-MCM-48 might allow for the rapid diffusion of the propylene monomer to active sites in the mesopores. To clarify the characteristics of pps(ppout, ppin) outside and inside the mesopores of Ti-containing mesoporous silicas, the as-polymerized PP was extracted with boiling ODCB for 8 h and the obtained ppout was characterized by means of GPC, DSC and 13C NMR. As summarized in Table 2, the molecular weight distribution of ppout was broader (Mw/Mn >15) than that of conventional polypropylene prepared with Ziegler-Natta catalysts (typically, Mw/Mn - 4 - 6). The isotactic pentad fraction [mmmm], Tm and the enthalpy of fusion ( A H m ) ofPP ~ were 30 - 60%, 150 - 160~ and ca. 20 J g-l, respectively. Taking into account the high Tm (corresponding to [mmmm] ~ 92%) and the low A H m (corresponding to [mmmm] < 40%) [13], these results strongly suggest that the obtained ppout is a mixture of polypropylenes with various stereo-regularities. Table 2 Results of propylene polymerization with Ti-containing mesoporous silica/Al(i-C4H9)3 catalyst. Catalyst Activity ppout ppin (g-PP Ti_mol.l h.i) Tm Mw Mw/ [mmmm] Tm Mw Mw/ [mmmm] (~ xlO 3 Mn (%) (~ xlO 3 Mn (%) Ti-MCM-41 1,097 151.4 252 >50 32 154.7 84 4.3 38 Ti-MCM-48
1,448
155.5 140
15
49
156.2
51
16
36
Ti-SBA-15
1,186
158.6 212
23
59
159.9
32
11
62
Polymerization conditions: Catalyst =0.5 g, AI(i-CaH9)3/Ti =2, Toluene =30 cm 3, Propylene =7 dm3(S.T.P.), Temp. =40~ Time =2 h. 1.3. Characterization of PP/Ti-containing mesoporous silica nanocomposite Next, following the removal of the polymer outside mesopores of Ti-containing mesoporous silica, the characterization of the system insoluble in boiling ODCB, the so-called PP/Ti-containing mesoporous silica nanocomposite, was carried out. Fig. 4 shows the XRD patterns of the Ti-SBA-15 after propylene polymerization and the PP/Ti-SBA-15 nanocomposite. The diffraction peaks assigned to m-form of PP were observed around 2 theta 15 - 25 ~ in the XRD pattern of the Ti-SBA-15 after polymerization although their intensities were very low. The retention of the Ti-SBA-15 characteristic diffraction peaks in the XRD pattern of the PP/Ti-SBA-15 nanocomposite indicates that the hexagonal framework structure
1441
(e)
(d) 10 tD
20
30
d
_
_
(b)
Y,
(a) ~ 1
I 2
I
I
I
I
I
3 4 5 6 2 theta (degree)
7
Fig. 4. XRD patterns of (a) T-SBA-15 after propylene polymerization and (b) PP/Ti-SBA-15 nanocomposite.
I
-100
L_
(a) !
-50
0 50 ppm Fig. 5.13C CP/MAS NMR spectra of (a) PP/Ti-MCM-41, (b) PP/Ti-MCM-48 (c) PP/Ti-SBA-15, (d) isotactic PP and (e) atactic PP.
ofTi-BA-15 is retained during and after the propylene polymerization. Fig. 5 shows the 13C CP/MAS NMR spectra of various PP/Ti-containing mesoporous silica nanocomposites. For a reference, the spectra of isotactic PP(ot-form, Mw=190,000, Mw/Mn=5.2, [mmmm]=99%) and atactic PP(Mw=350,000, Mw/Mn=4.2) are also shown. The spectrum of the PP/Ti-MCM-41 nanocomposite was similar to that of the conventional isotactic PP. Although the 13C CP/MAS NMR spectra of the PP/Ti-MCM-48 and the PP/Ti-SBA-15 were slightly different from those of isotactic and atactic PPs, these results indicate that the PP/Ti-containing mesoporous silica nanocomposites consist of Ti-containing mesoporous silica and PP confined in mesopores. Furthermore, taking into account the fact that the PP confined in mesopores of Ti-mesoporous silica could not be extracted with boiling ODCB, there seems to exist strong interactions between the confined PP and the pore wall. Fig. 6 shows the N2 adsorption isotherms on the Ti-SBA-15 and the PP/Ti-SBA-15 nanocomposite. For the PP/Ti-SBA-15, a slight decrease in the amount of N2 adsorbed was observed due to the existence of PP in the mesopores. A quantitative analysis of the content of PP in the PP/Ti-containing mesoporous silica nanocomposites was performed with thermogravimetric analysis (TG). The weight loss and exothermic peaks between 200 and 500~ are assigned to the decomposition of organic compounds, that is, PP (Fig. 7). As expected, the weight loss between 200 and 500~ decreased considerably after the ODCB extraction, which removes ppout. The remaining PPs correspond to those in the PP/Ti-MCM-41, the PP/Ti-MCM-48 and the PP/Ti-SBA-15 nanocomposites were ca.15, 8.9 and 8.6 wt%. Assuming that the reduction in the pore volume of Ti-containing mesoporous silica measured by nitrogen adsorption was due only to the inclusion of PP and the density of amorphous PP is 0.85 g m1-1 [14], the contents of PPs confined in the PP/Ti-MCM-41, the PP/Ti-MCM-48 and the PP/SBA-15 nanocomposites are calculated to be 13.9, 8.6 and 8.5 wt%, respectively. These contents of PPs are comparable to those estimated by TG. Namely, only 10 - 30 vol% of total pore volume is occupied by PP. This is a significant effect of the
1442 600 (a) 500
A
_
o.o
AA
A
a; 400
o
oog~~
c
o
L
A
(b)
o
,-t::l cD ,.Q O
Z
A
300
(a)
zx8 o
200 r-4
(c)
~
000
O0
1000 t I
0.0
0.2
I
I
0.4 0.6 P/P0
I
0.8
I
1.0
Fig. 6. N2 adsorption isotherms of(a) Ti-SBA-15 and (b) PP/Ti-SBA- 15 nanocomposite. The pore volume of Ti-SBA-15 in the PP/Ti-SBA- 15 nanocomposite was normalized based on 1 g of Ti-SBA-15.
0
I
I
I
200 400 Temperature (~
I
I
i
600
800
Fig. 7. TG curves of (a) PP/Ti-MCM-41, (b) PP/Ti-MCM-48 and (c) PP/Ti-SBA- 15 nanocomposites.
confined geometry, and we here with tentatively attributed it to the formation of PP near the ends of mesopores, which can cause a retardation of propylene monomer diffusion to the active sites deeper inside the mesopores, and thus result in such low content of confined polypropylene in the mesopores.
3.4. C h a r a c t e r i z a t i o n
of confined polypropylene
To characterize the confined PPs in the mesopores of the Ti-MCM-41, the Ti-MCM-48 and the Ti-SBA-15, the PP/Ti-containing mesoporous silica nanocomposites were treated with a 12M NaOH aqueous solution at 120~ for 48 h, and this was followed by extraction with ODCB at 140~ As shown in Fig. 8 and Table :2, the average molecular weight (Mw) of ppin was much smaller and its distribution was .remarkably sharper than those of ppout, except for Ti-MCM-48. The narrower Mw/Mn of ppm might originate from the smaller Mw of ppm as compared with Mw of ppout. The isotacticity pentad fraction [mmmm] of ppin was 38 - 6 2 % , as low as that ofpp~ - 59%). As the alkali treatment does not significantly decompose the polypropylene molecules, therefore, these facts indicate that the behavior of propylene polymerization is strongly affected by the confined space of the mesopores.
1443 1.0
(c)
(B)
(A) 0.8 0.6 O
(b) ~ / ~ x , ,
0.4 0.2
ss~ SS
~
0.0
l/
2
3
4 5 Log Mw
6
72
3
I
!
I
4 5 Log Mw
6
72
3
4 5 6 Log Mw
7
Fig. 8. GPC curves of(a) ppout and (b) ppin of(A) Ti-MCM-41, (B) Ti-MCM-48 and (C) Ti-SBA-15. The DSC curves of PPs in are shown in Fig. (c) 9. For a reference, the DSC curves of the PP/Ti-containing mesoporous silica nanocomposites are also showed. After ppin k...._-was extracted from the mesopores, a melting point Tm was clearly observed at 150 - 160~ whereas the same polymer when located . . . . . . . . . . . . . . . . . . . . . ,,., , . . . - . within the mesopores showed no Tin. This is direct evidence that the severe confinement of the mesopore geometries hinders the (a) polypropylene crystallization. This behavior can be attributed to the existence of bound PP physisorbed on the inner-wall of Ti-containing mesoporous silicas, analogous to the one that Tsagaropoulos and Eisenberg ! ! ! ! reported for polymer chains filled with very 150 200 -50 0 50 100 fine silica particles of high surface area [15]. Temperature (~ The strong polymer/inorganic interactions can restrict the mobility of the bound polymer Fig. 9. DSC curves of (__)ppin and chains and prevent the polymer from (.... )PP/Ti-containing mesoporous silica. crystallizing. Another possible explanation is (a) Ti-MCM-41, (b) Ti- MCM-48 and (c) the purely geometric effect of the Ti-SBA-15. confinement. Vaia et al. attributed the loss of crystallinity of poly(ethylene oxide) (PEO) to the severe confinement of the polymer between silicate layers in the intercalated PEO/montmorillonite nanocomposite [ 16]. Okamoto et al. argued that it might be difficult for intercalated polypropylene chains in PP/clay nanocomposites to crystallize because of the limited space between clay layers (2 - 3 nm) [17]. Manias et al. also reported by using computer simulations that the intercalated polymer chains in PEO/clay nanocomposite adopt a disordered, liquid-like structure with no crystallinity or periodic ordering of the C-C-O bonds L
. . m - .
,--" m
a
w
----'4-
1444 [ 18]. Although at the present time we can not give definitive argumentation for the molecular mechanisms behind the disappearance of the PP melting point in the confined PP in the Ti-containing mesoporous silica, we tentatively attribute this behavior to the same confinements as those in the two-dimensional confinement reported so far. 4. C O N C L U S I O N S Ti-MCM-48 and Ti-SBA-15 as well as Ti-MCM-41 also exhibited the isotactic polymerization of propylene. Despite being polymerized concurrently, polypropylenes outside and insides the mesopores of Ti-containing mesoporous silicas had different characteristics. The Mw of ppin was much smaller and its distribution (Mw/Mn) was remarkably sharper than those of ppout. The isotacticity [mmmm] of PP prepared by Ti-SBA-15 was larger than that prepared by Ti-MCM-41. The severe confinement of the mesopore geometry hindered the polypropylene crystallization. From all above results, it was concluded that the behavior of propylene polymerization is strongly affected by the pore size and structure of the mesopores of Ti-containing mesoporous silicas used. REFERENCES
[1] [2] [3] [4] [5] [6] [7]
Y.S. Ko, T.K. Han, J.W. Park and S.I. Woo, Macromol. Rapid Commun., 17 (1996) 749. K. Kageyama, J.L. Tamazawa and T. Aida, Science, 285 (1999) 2113. J. He, X. Duan, D.G. Evans and R.F. Howe, J. Porous Mater., 9 (2002) 49. P.T. Tanev, M. Chibwe and T.J. Pinnavaia, Nature, 368 (1994) 321. A. Corma, M.T. Navarro and J.P. Pariente, J. Chem. Soc., Chem. Comm., (1994) 147. T. Tatsumi, K.A. Koyano and N. Igarashi, J. Chem.Soc., Chem. Commun., (1998) 325. Y. Oumi, A. Hanai, T. Miyazaki, H. Nakajima, S. Hosoda, T. Teranishi and T. Sano, Stud. Surf. Sci. Catal., 146 (2003) 753. [8] H. Nakajima, K. Yamada, Y. Iseki, S. Hosoda, A. Hanai, Y. Oumi, T. Teranishi and T. Sano, J. Polym. Sci. Part B: Polym. Phys., 41 (2003) 3324. [9] J.M. Kim, J.H. Kwak, S. Jun and R. Ryoo, J. Phys. Chem., 99 (1995) 16742. [10] R. Ryoo, S.H. Joo and J.M. Kim, J. Phys. Chem. B, 103 (1999) 7435. [ 11] D. Zhao, J. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka and G.D. Stucky, Science, 279 (1998) 548. [12] L. Marchese, T. Maschmeyer, E. Gianotti, S. Coluccia and J.M. Thomas, J. Phys. Chem. B, 101 (1997) 8836. [ 13 ] E. Martuscelli, M. Pracella and A. Zambelli, J. Poly. Sci. Polym. Phys. Ed., 18 (1980) 619. [ 14] B. Wunderlich, Macromolecular Physics, Academic, New York, 1980. [15] G. Tsagaropoulos andA. Eisenberg, Macromolecules, 28 (1995) 6067. [16] R.A. Vaia, B.B. Sauer, O.K. Tse and E.P. Giannelis, J. Polym. Sci. Part B: Polym. Phys., 35 (1997) 59. [17] P. Maiti, P.H. Nam, M. Okamoto, H. Hasegawa and A. Usuki, Macromolecules, 35 (2002) 2042. [ 18] V. Kuppa, S. Menakanit, R. Krishnamoorti and E. Manias, J. Polym. Sci. Part B: Polym. Phys.,. 41 (2003) 3285.
1437
Propylene polymerization behavior of Ti-containing mesoporous silicas Y. O u m i a, S. T a k a s h i m a a, A. Hanai a, H. N a k a j i m a b, K. Y a m a d a b, S. Hosoda b and T. Sano a
aSchool of Materials Science, Japan Advanced Institute of Science and Technology, Tatsunokuchi, Ishikawa 923-1292, Japan ; E-mail: [email protected] bpetrochemicals Research Laboratory, Sumitomo Chemical Co. Ltd., Sodegaura, Chiba 299-0295, Japan Ti-containing mesoporous silicas (Ti-MCM-41, Ti-MCM-48 and Ti-SBA-15) were prepared by the post-synthesis with Ti(OC4H9)4 and their propylene polymerization behavior was investigated. It was found that these Ti-containing mesoporous silicas combined with Al(i-C4H9)3 provide the isotactic polypropylene. Polypropylenes outside and inside the mesopores of Ti-containing mesoporous silicas had different characteristics despite being polymerized concurrently. The crystallization of polypropylenes confined in the mesopores was considerably depressed due to the limited space. 1. INTRODUCTION Recently, there has been a great deal of interest in the behavior of polymers under confined geometries for understanding and controlling the properties of polymers on micrometer and nanometer scales. Several inorganic porous materials such as layered silicate, mesoporous silica and porous glass are used as a nano-reactor. However, there are only a few papers concerning polymerization of olefins such as ethylene and propylene with mesoporous materials, in which organometallic complexes are impregnated or grafted on the surfaces of mesopores [ 1-3]. Since the discover of ordered mesoporous silicas such as MCM-41, MCM-48 and SBA-15, much effort has been paid to incorporation of various metals into these framework structures by the direct hydrothermal synthesis and the post-synthesis methods for heterogeneous catalysis and adsorption. It is well known that Ti- and V-containing mesoporous silicas are active for selective oxidation of a wide variety of organic substrates with the environmentally friendly oxidant [4-6]. From such viewpoints, we have studied the olefin polymerization in mesopores, and very recently found metal-containing MCM-41, especially Ti-containing MCM-41 combined with alkylaluminums can provide isotactic polypropylene with broad molecular weight distribution [7,8]. However, the detailed olefin polymerization behavior in mesopores is still an open question. In this paper, therefore, several Ti-containing mesoporous silicas with different mesoporous structures such as Ti-MCM-41, Ti-MCM-48 and Ti-SBA-15 were prepared and used to polymerize propylene. By analyzing the characteristics of polypropylenes both outside and inside the mesopores, the influences of the mesopore structure of Ti-containing mesoporous silica on the polymerization behavior was investigated in detail.
1438 2. EXPERIMENTAL
2.1. Preparation and characterization of Ti-containing mesoporous silicas Ti-containing mesoporous silicas were prepared by the post-synthesis method. The parent siliceous MCM-41, MCM-48 and SBA-15 were prepared following the procedures described in the literature [9-11 ]. The siliceous mesoporous silicas were calcined at 500~ for 10 h to decompose the surfactant. 1 g of the mesoporous silica calcined at 500~ for 10 h was dispersed in 10 ml of dry toluene containing 3 mmol of Ti(OC4H9)4. The mixture was refluxed at 115~ for 24 h. The product was filtered, washed with dry toluene several times, dried at room temperature and then calcined at 500~ in air for 5 h. Identification of the products was carried out by X-ray diffraction (XRD, Rigaku RINT-2000). Elemental composition was determined by X-ray fluorescence (XRF, Philips PW-2400). Textural properties (BET surface area, pore diameter, pore volume) were evaluated by nitrogen adsorption at -196~ (Bel Japan, Belsorp 28SA). UV-Vis spectra were recorded on a Hitachi U-3310 spectrometer. 2.2. Polymerization of propylene Propylene polymerization was conducted at 40~ in a 100 cm 3 stainless steel autoclave equipped with a magnetic stirrer. After the reactor was filled with nitrogen, measured amounts of Ti-containing mesoporous silica evacuated at 400~ for 8 h, toluene and Al(i-C4H9)3 were added to the reactor in this order and the mixture was kept standing at room temperature for 15 min. The reactor was evacuated at liquid nitrogen temperature, and the 7 dm 3 of propylene was introduced. Polymerization was started by quickly heating the reactor up to the polymerization temperature (40~ The polymerization reaction was terminated by adding acidified methanol (5 wt% HC1). For characterization of the polypropylene outside mesopores of Ti-containing mesoporous silica (ppout), the as-synthesized polypropylene was extracted with the boiling o-dichlorobenzene (ODCB). Also, to obtain the polypropylene inside mesopores (pp,n), the
t
Extraction with ~ bc ling ODCB Y
(
(d)
Alkali treatment
Extraction wit boiling ODCB Fig. 1. Schematic diagram of various treatments. (a) as-polymerized PP, (b) PP/Ti-containing mesoporous silica nanocomposite, (c) PP outside mesopores of Ti-containing mesoporous silica and (d) PP inside mesopores ofTi-containing mesoporous silica.
1439 PP/Ti-containing mesoporous silica nanocomposite was treated with a 12M NaOH aqueous solution at 120~ for 48 h and this was followed by extraction with ODCB (Fig. 1). The weight-average molecular weight (Mw) and molar mass distribution (Mw/Mn, Mn: number-average molecular weight) of the polymers were measured at 145~ by gel-permeation chromatography (GPC, Senshu Scientific SSC7100) using ODCB as a solvent. The melting points (Tm) of the polymers were measured on a Seiko DSC-220C calorimeter with a heating rate of 10~ min -1 in the temperature range of-40 - 200~ ~3C NMR spectra of the polymers were measured in 1,2,4-trichloro benzene/benzene-d6 (9/1 v/v) at 140~ using a Varian GEM-300 spectrometer operating at 75.4 MHz. Solid state 13C CP/MAS NMR spectra of the PP/Ti-containing mesoporous silicas were recorded using a zirconia rotor with a 7 mm diameter on a Varian VXP-400 spectrometer at 100.7 MHz. 3. RESULTS AND DISCUSSION 3.1. Preparation
of Ti-containing
mesoporous
silicas
The parent siliceous MCM-41, MCM-48 and SBA-15 exhibited the typical XRD patterns. MCM-41 and SBA- 15 exhibited three diffraction peaks ((100), (110), (200)), indicating the hexagonal framework structure. On the other hand, MCM-48 exhibited the diffraction peaks corresponding to the highly branched and interwoven three-dimensional channel system. Fig. 2 shows the XRD patterns of Ti-MCM-41, Ti-MCM-48 and Ti-SBA-15 prepared by the post-synthesis method. All samples gave slightly lower quality XRD patterns than those of the corresponding parents. The XRD patterns were found to be nearly free from crystalline TiO2.
(•i
(lOO) lo)2 oo)
(c)
.(211)
r~
~0~420)~(322)
~D
(b)
100)
0) i
0
I
_
(a)
4 6 2 theta (degree) Fig. 2. XRD patterns of (a) Ti-MCM-41, (b) Ti-MCM-48 and (c) Ti-SBA-15.
I
!
I
2
200
,
I
,
I
,
I
,
I
,
300
400 500 600 700 800 Wavelength (nm) Fig. 3. UV-Vis spectrum of Ti-MCM-48.
Table 1 Characteristics of various Ti-containing mesoporous silicas. Sample (Si/Ti ratio) BET surface area (m 2 g-l) Pore volume (cm 3 g-l) Pore diameter (nm) Ti-MCM-41 (22.4) Ti-MCM-48 (21.7) Ti-SBA-15 (22.5)
908 1,215 647
0.72 1.10 0.72
2.74 2.52 5.80
1440 This was also confirmed from UV-Vis analysis. As can be seen in Fig. 3, the UV-Vis spectrum showed an absorption band at ca. 220 nm due to a charge transfer between framework oxygen and tetrahedral Ti(IV) [12]. However, the spectrum also showed a shoulder around 330 nm corresponding to octahedrally coordinated Ti species, indicating the presence of trace amount of octahedral TiO2. Characteristics of these Ti-containing mesoporous silicas are listed in Table 1. 3.2. Polymerization of propylene Propylene polymerization was carried out using the Ti-containing mesoporous silicas prepared. All of the Ti-containing mesoporous silicas combined with A1 (i-C4H9)3 displayed activity and gave selectively isotactic polypropylene (PP). The order of the polymerization activity was as follows: Ti-MCM-48>Ti-SBA-15>Ti-MCM-41 (Table 2). This suggests that the pore size and structure of mesoporous silica affect the polymerization activity; three-dimensional structure of Ti-MCM-48 might allow for the rapid diffusion of the propylene monomer to active sites in the mesopores. To clarify the characteristics of pps(ppout, ppin) outside and inside the mesopores of Ti-containing mesoporous silicas, the as-polymerized PP was extracted with boiling ODCB for 8 h and the obtained ppout was characterized by means of GPC, DSC and 13C NMR. As summarized in Table 2, the molecular weight distribution of ppout was broader (Mw/Mn >15) than that of conventional polypropylene prepared with Ziegler-Natta catalysts (typically, Mw/Mn - 4 - 6). The isotactic pentad fraction [mmmm], Tm and the enthalpy of fusion ( A H m ) ofPP ~ were 30 - 60%, 150 - 160~ and ca. 20 J g-l, respectively. Taking into account the high Tm (corresponding to [mmmm] ~ 92%) and the low A H m (corresponding to [mmmm] < 40%) [13], these results strongly suggest that the obtained ppout is a mixture of polypropylenes with various stereo-regularities. Table 2 Results of propylene polymerization with Ti-containing mesoporous silica/Al(i-C4H9)3 catalyst. Catalyst Activity ppout ppin (g-PP Ti_mol.l h.i) Tm Mw Mw/ [mmmm] Tm Mw Mw/ [mmmm] (~ xlO 3 Mn (%) (~ xlO 3 Mn (%) Ti-MCM-41 1,097 151.4 252 >50 32 154.7 84 4.3 38 Ti-MCM-48
1,448
155.5 140
15
49
156.2
51
16
36
Ti-SBA-15
1,186
158.6 212
23
59
159.9
32
11
62
Polymerization conditions: Catalyst =0.5 g, AI(i-CaH9)3/Ti =2, Toluene =30 cm 3, Propylene =7 dm3(S.T.P.), Temp. =40~ Time =2 h. 1.3. Characterization of PP/Ti-containing mesoporous silica nanocomposite Next, following the removal of the polymer outside mesopores of Ti-containing mesoporous silica, the characterization of the system insoluble in boiling ODCB, the so-called PP/Ti-containing mesoporous silica nanocomposite, was carried out. Fig. 4 shows the XRD patterns of the Ti-SBA-15 after propylene polymerization and the PP/Ti-SBA-15 nanocomposite. The diffraction peaks assigned to m-form of PP were observed around 2 theta 15 - 25 ~ in the XRD pattern of the Ti-SBA-15 after polymerization although their intensities were very low. The retention of the Ti-SBA-15 characteristic diffraction peaks in the XRD pattern of the PP/Ti-SBA-15 nanocomposite indicates that the hexagonal framework structure
1441
(e)
(d) 10 tD
20
30
d
_
_
(b)
Y,
(a) ~ 1
I 2
I
I
I
I
I
3 4 5 6 2 theta (degree)
7
Fig. 4. XRD patterns of (a) T-SBA-15 after propylene polymerization and (b) PP/Ti-SBA-15 nanocomposite.
I
-100
L_
(a) !
-50
0 50 ppm Fig. 5.13C CP/MAS NMR spectra of (a) PP/Ti-MCM-41, (b) PP/Ti-MCM-48 (c) PP/Ti-SBA-15, (d) isotactic PP and (e) atactic PP.
ofTi-BA-15 is retained during and after the propylene polymerization. Fig. 5 shows the 13C CP/MAS NMR spectra of various PP/Ti-containing mesoporous silica nanocomposites. For a reference, the spectra of isotactic PP(ot-form, Mw=190,000, Mw/Mn=5.2, [mmmm]=99%) and atactic PP(Mw=350,000, Mw/Mn=4.2) are also shown. The spectrum of the PP/Ti-MCM-41 nanocomposite was similar to that of the conventional isotactic PP. Although the 13C CP/MAS NMR spectra of the PP/Ti-MCM-48 and the PP/Ti-SBA-15 were slightly different from those of isotactic and atactic PPs, these results indicate that the PP/Ti-containing mesoporous silica nanocomposites consist of Ti-containing mesoporous silica and PP confined in mesopores. Furthermore, taking into account the fact that the PP confined in mesopores of Ti-mesoporous silica could not be extracted with boiling ODCB, there seems to exist strong interactions between the confined PP and the pore wall. Fig. 6 shows the N2 adsorption isotherms on the Ti-SBA-15 and the PP/Ti-SBA-15 nanocomposite. For the PP/Ti-SBA-15, a slight decrease in the amount of N2 adsorbed was observed due to the existence of PP in the mesopores. A quantitative analysis of the content of PP in the PP/Ti-containing mesoporous silica nanocomposites was performed with thermogravimetric analysis (TG). The weight loss and exothermic peaks between 200 and 500~ are assigned to the decomposition of organic compounds, that is, PP (Fig. 7). As expected, the weight loss between 200 and 500~ decreased considerably after the ODCB extraction, which removes ppout. The remaining PPs correspond to those in the PP/Ti-MCM-41, the PP/Ti-MCM-48 and the PP/Ti-SBA-15 nanocomposites were ca.15, 8.9 and 8.6 wt%. Assuming that the reduction in the pore volume of Ti-containing mesoporous silica measured by nitrogen adsorption was due only to the inclusion of PP and the density of amorphous PP is 0.85 g m1-1 [14], the contents of PPs confined in the PP/Ti-MCM-41, the PP/Ti-MCM-48 and the PP/SBA-15 nanocomposites are calculated to be 13.9, 8.6 and 8.5 wt%, respectively. These contents of PPs are comparable to those estimated by TG. Namely, only 10 - 30 vol% of total pore volume is occupied by PP. This is a significant effect of the
1442 600 (a) 500
A
_
o.o
AA
A
a; 400
o
oog~~
c
o
L
A
(b)
o
,-t::l cD ,.Q O
Z
A
300
(a)
zx8 o
200 r-4
(c)
~
000
O0
1000 t I
0.0
0.2
I
I
0.4 0.6 P/P0
I
0.8
I
1.0
Fig. 6. N2 adsorption isotherms of(a) Ti-SBA-15 and (b) PP/Ti-SBA- 15 nanocomposite. The pore volume of Ti-SBA-15 in the PP/Ti-SBA- 15 nanocomposite was normalized based on 1 g of Ti-SBA-15.
0
I
I
I
200 400 Temperature (~
I
I
i
600
800
Fig. 7. TG curves of (a) PP/Ti-MCM-41, (b) PP/Ti-MCM-48 and (c) PP/Ti-SBA- 15 nanocomposites.
confined geometry, and we here with tentatively attributed it to the formation of PP near the ends of mesopores, which can cause a retardation of propylene monomer diffusion to the active sites deeper inside the mesopores, and thus result in such low content of confined polypropylene in the mesopores.
3.4. C h a r a c t e r i z a t i o n
of confined polypropylene
To characterize the confined PPs in the mesopores of the Ti-MCM-41, the Ti-MCM-48 and the Ti-SBA-15, the PP/Ti-containing mesoporous silica nanocomposites were treated with a 12M NaOH aqueous solution at 120~ for 48 h, and this was followed by extraction with ODCB at 140~ As shown in Fig. 8 and Table :2, the average molecular weight (Mw) of ppin was much smaller and its distribution was .remarkably sharper than those of ppout, except for Ti-MCM-48. The narrower Mw/Mn of ppm might originate from the smaller Mw of ppm as compared with Mw of ppout. The isotacticity pentad fraction [mmmm] of ppin was 38 - 6 2 % , as low as that ofpp~ - 59%). As the alkali treatment does not significantly decompose the polypropylene molecules, therefore, these facts indicate that the behavior of propylene polymerization is strongly affected by the confined space of the mesopores.
1443 1.0
(c)
(B)
(A) 0.8 0.6 O
(b) ~ / ~ x , ,
0.4 0.2
ss~ SS
~
0.0
l/
2
3
4 5 Log Mw
6
72
3
I
!
I
4 5 Log Mw
6
72
3
4 5 6 Log Mw
7
Fig. 8. GPC curves of(a) ppout and (b) ppin of(A) Ti-MCM-41, (B) Ti-MCM-48 and (C) Ti-SBA-15. The DSC curves of PPs in are shown in Fig. (c) 9. For a reference, the DSC curves of the PP/Ti-containing mesoporous silica nanocomposites are also showed. After ppin k...._-was extracted from the mesopores, a melting point Tm was clearly observed at 150 - 160~ whereas the same polymer when located . . . . . . . . . . . . . . . . . . . . . ,,., , . . . - . within the mesopores showed no Tin. This is direct evidence that the severe confinement of the mesopore geometries hinders the (a) polypropylene crystallization. This behavior can be attributed to the existence of bound PP physisorbed on the inner-wall of Ti-containing mesoporous silicas, analogous to the one that Tsagaropoulos and Eisenberg ! ! ! ! reported for polymer chains filled with very 150 200 -50 0 50 100 fine silica particles of high surface area [15]. Temperature (~ The strong polymer/inorganic interactions can restrict the mobility of the bound polymer Fig. 9. DSC curves of (__)ppin and chains and prevent the polymer from (.... )PP/Ti-containing mesoporous silica. crystallizing. Another possible explanation is (a) Ti-MCM-41, (b) Ti- MCM-48 and (c) the purely geometric effect of the Ti-SBA-15. confinement. Vaia et al. attributed the loss of crystallinity of poly(ethylene oxide) (PEO) to the severe confinement of the polymer between silicate layers in the intercalated PEO/montmorillonite nanocomposite [ 16]. Okamoto et al. argued that it might be difficult for intercalated polypropylene chains in PP/clay nanocomposites to crystallize because of the limited space between clay layers (2 - 3 nm) [17]. Manias et al. also reported by using computer simulations that the intercalated polymer chains in PEO/clay nanocomposite adopt a disordered, liquid-like structure with no crystallinity or periodic ordering of the C-C-O bonds L
. . m - .
,--" m
a
w
----'4-
1444 [ 18]. Although at the present time we can not give definitive argumentation for the molecular mechanisms behind the disappearance of the PP melting point in the confined PP in the Ti-containing mesoporous silica, we tentatively attribute this behavior to the same confinements as those in the two-dimensional confinement reported so far. 4. C O N C L U S I O N S Ti-MCM-48 and Ti-SBA-15 as well as Ti-MCM-41 also exhibited the isotactic polymerization of propylene. Despite being polymerized concurrently, polypropylenes outside and insides the mesopores of Ti-containing mesoporous silicas had different characteristics. The Mw of ppin was much smaller and its distribution (Mw/Mn) was remarkably sharper than those of ppout. The isotacticity [mmmm] of PP prepared by Ti-SBA-15 was larger than that prepared by Ti-MCM-41. The severe confinement of the mesopore geometry hindered the polypropylene crystallization. From all above results, it was concluded that the behavior of propylene polymerization is strongly affected by the pore size and structure of the mesopores of Ti-containing mesoporous silicas used. REFERENCES
[1] [2] [3] [4] [5] [6] [7]
Y.S. Ko, T.K. Han, J.W. Park and S.I. Woo, Macromol. Rapid Commun., 17 (1996) 749. K. Kageyama, J.L. Tamazawa and T. Aida, Science, 285 (1999) 2113. J. He, X. Duan, D.G. Evans and R.F. Howe, J. Porous Mater., 9 (2002) 49. P.T. Tanev, M. Chibwe and T.J. Pinnavaia, Nature, 368 (1994) 321. A. Corma, M.T. Navarro and J.P. Pariente, J. Chem. Soc., Chem. Comm., (1994) 147. T. Tatsumi, K.A. Koyano and N. Igarashi, J. Chem.Soc., Chem. Commun., (1998) 325. Y. Oumi, A. Hanai, T. Miyazaki, H. Nakajima, S. Hosoda, T. Teranishi and T. Sano, Stud. Surf. Sci. Catal., 146 (2003) 753. [8] H. Nakajima, K. Yamada, Y. Iseki, S. Hosoda, A. Hanai, Y. Oumi, T. Teranishi and T. Sano, J. Polym. Sci. Part B: Polym. Phys., 41 (2003) 3324. [9] J.M. Kim, J.H. Kwak, S. Jun and R. Ryoo, J. Phys. Chem., 99 (1995) 16742. [10] R. Ryoo, S.H. Joo and J.M. Kim, J. Phys. Chem. B, 103 (1999) 7435. [ 11] D. Zhao, J. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka and G.D. Stucky, Science, 279 (1998) 548. [12] L. Marchese, T. Maschmeyer, E. Gianotti, S. Coluccia and J.M. Thomas, J. Phys. Chem. B, 101 (1997) 8836. [ 13 ] E. Martuscelli, M. Pracella and A. Zambelli, J. Poly. Sci. Polym. Phys. Ed., 18 (1980) 619. [ 14] B. Wunderlich, Macromolecular Physics, Academic, New York, 1980. [15] G. Tsagaropoulos andA. Eisenberg, Macromolecules, 28 (1995) 6067. [16] R.A. Vaia, B.B. Sauer, O.K. Tse and E.P. Giannelis, J. Polym. Sci. Part B: Polym. Phys., 35 (1997) 59. [17] P. Maiti, P.H. Nam, M. Okamoto, H. Hasegawa and A. Usuki, Macromolecules, 35 (2002) 2042. [ 18] V. Kuppa, S. Menakanit, R. Krishnamoorti and E. Manias, J. Polym. Sci. Part B: Polym. Phys.,. 41 (2003) 3285.