DIBUTYLTIN DILAURATE-CATALYZED DIISOCYANATE COUPLING- GRAFTING OF POLY(ETHYLENE GLYCOL) ONTO NANO-SI

Chinese Journal of Polymer Science Vol. 23, No.4, (2005), 379- -385Chinese Journal ofPolymer ScienceC2005 World-ScientificDIBUTYLTIN DILAURATE-CATALYZED DIISOCYANATE COUPLING-GRAFTING OF POLY(ETHYLENE GLYCOL) ONTO NANO-SILICA SURFACETHROUGH A ONE-STEP PROCEDUREBin Wang, Jian Yu and Zhao-xia Guo'Institute of Polymer Science and Engineering, Department of Chemical Engineering, School of Materials Science andEngineering, Tsinghua University, Bejing 100084, ChinaAbstract Grafting polymer glycols onto nano-silica surface through one-step procedure was investigated. The majorcharacteristic of this procedure is that all the materials and reagents (silica, PEG, TDI, DBTDL, solvent) required for graftingwere added simultaneously into the reaction vessel. TDI and DBTDL were used as coupling agent and catalyst, respectively.The products were characterized by FTIR, TGA, elemental analyses and TEM, giving evidence for successful grafting ofPEG. Possible mechanism of this grafing was studied and two grating processes were proposed. The process through whichthe grafting proceeds depends on the reaction temperature. Effect of molecular weight of PEG on grating was also investigated.Keywords: Silica; PEG; Toluene disocyanate; Grafting.INTRODUCTIONIn recent years, a considerable amount of research deals with inorganic nanoparticle/polymer composites whichare expected to exhibit excellent mechanical performance and diverse functions such as UV-resisting and anti-bacterial properties. The major annoying problems associated with this type of composites are the dispersibilityof inorganic particle and their incompatibility with the polymer matrix. Surface modification on inorganicnanoparticles is an effective way to solve the above-mentioned problems. Although physical methods, in whichno covalent interaction is involved between inorganic nanoparticle and organic modifier, are simple, chemicalmodifications are preferred in view of the firmness of the layers between the two phases. The desorption at theseinterfacial layers may 0ccur under high shearing of processing or in solvents during subsequent usage, if there isno covalent bonding between inorganic nanoparticles and the organic modifier.As an effective chemical modification method, grafting of polymers onto inorganic surface is of growinginterest. A number of works have been published, involving mainly silica. The procedures can be divided intothree types: (1) grafting-from approach! , featuring the introduction of compounds bearing initiating groupssuch as azo and peroxide onto silica surface, followed by initiation of polymerization reaction by those groups,(2) grafting- through approach5s 61, featuring the introduction of compounds bearing polymerizable double bondonto silica surface, followed by copolymerization with vinyl polymers and (3) grafing-onto approach!'featuring the introduction of compounds bearing functional groups such as amine and isocyanate group ontosilica surface, followed by reaction with polymers bearing appropriate functional groups; or functionalization ofcertain polymer first, followed by reaction with the hydroxyl group of silica.In our present work, a one-step procedure is achieved in which all the reactants, solvent and catalyst wereintroduced to the reaction vessel together for grafting PEG onto nano-silica surface using TDI as a couplingagent. DBTDL, a well-known powerful catalyst for synthesis of polyurethane, was used to accelerate the graftingreaction. The mechanism of the grafting reaction was studied at中国煤化工"Corresponding author: Zhao-xia Guo(郭朝霞), E mail: guozx@ mail.tsilTYHCNMHGReceived June 18, 2004; Revised August 1, 2004; Accepted August 11, 2004380B. Wang et al.EXPERIMENTALMaterials and ReagentsNano-silica (type MN1S) was obtained from Zhoushan-Mingri Company (China) which has a mean particle sizeof 15 nm and a specific area of 160 m2/g. PEG with various molecular weights (400, 1000, 2000 and 6000) wereobtained from T.T.R.C (China). Both silica and PEG were dried in vacum oven at 110°C for 7 h before use.TDI (80/20 mixture of 2,4- and 2,6-isomers) and dibutyltin dilaurate (DBTDL) were used as received. Toluenewas dried over 4 A molecular sieves.One-step Grafting ProcedureIn a typical procedure, a mixture of silica (1 g), PEG2000 (1.4 g, 0.7 mmol), TDI (0.244 g, 1.4 mmol), andDBTDL (0.088 g, 0.14 mmol) was heated to and held at 80°C for 7 h under nitrogen atmosphere. After cooling,methanol (10 mL) was added to quench unreacted isocyanate groups. The product was then centrifuged andwashed with toluene for three times, then extracted in a Soxhlet apparatus with toluene for 24 h and subsequentlywith methanol for another 24 h to completely remove any ungrafted PEG and other possible derivatives.The change of reaction conditions and procedure for relevant experiments are specified in the relatedcontents of this paper.CharacterizationFTIR spectra of nano-silica and modified silica were recorded on a Nicolet 560 FTIR spectrometer.Thermogravimetric analysis (TGA) was carried out on a TA 2050 instrument with a heating rate of 20 K/min inflowing N2. Elemental analyses (EA) were performed on an Elementar Vario EL analyzer. TEM photographswere obtained using a Hitachi H800 eletron microscope.The percentages of graftings were calculated from TGA curves. As it will be explained later, the weight lossbetween 150°C and 700°C represents the weight of overall grafted organic substance. Among the total weightloss, weight loss between 150°C and 300°C mainly derives from TDI grafting, and weight loss between 300°Cand 700°C mainly derives from PEG grafting. The overall grafting, TDI grafting and PEG grafting werecalculated by the following equations:Overall grafting(%)=Overall organic compsition(g)x100(1)Bare silica(g)TDI gafting(%)=Grafed TDI(g)100(2)PEG gafting(%)=Grafted PEG(g)x 100(3)RESULTS AND DISCUSSIONEvidence of GraftingGrafting of PEG onto nano-silica was confirmed by FTIR, TGA, EA and TEM. As shown in Fig. 1, FTIRspectrum of the product shows characteristic stretching vibration of aromatic C=C at 1608 cm~' and CH2stretching vibration at 2954, 2916 and 2881 cmi , indicating the existence of both TDI and PEG on silica surface.The presence of urethane and amide bands at 1643 and 1541 cm~' reveal the formation of urethane bondsbetween TDI and silica, and perhaps also between TDI and PEG. Furthermore, the intensity of the OH peak at3458 cm~ 'decreases significantly (taking the peak at 471 cm !I as reference peak) and the NH (of the urethanestructure) peak appears at 3296 cm , proving the transformation of silanol group into urethane structure.中国煤化工YHCNMHGGrafing of PEG onto Nano-silica381105入<3458s-(291616433296 .s5-300020001000s00400800 100Wavenumber (cm~)Temperature (C)Fig. 1 FTIR spetra of original silica (top line)Fig. 2 TGA curves of original silica (1), TDI graftedand modifed silica (bottom line)silica (2), product obtained from the one-step procedure3), control experiment with PEG (4) and controlexperiment with methanol-capped TDI-PEG-TDI(5)There are three possibilities in which PEG would be grafted onto silica surface: (1) grafting of PEG throughTDI coupling, as expected; (2) grafting of PEG through ester formation between hydroxyl group of PEG andsilica silanol group; (3) firm adsorption through hydrogen bonding between etheric oxygen of PEG and silanolgroup of silica. The latter two are well known in some situations. In order to find out whether and to what extentthese interactions occur in our case, two control experiments were carried out. In one experiment, silica wastreated with PEG under the same conditions as in the typical procedure, 2% of PEG remains on silica surfaceafter Soxhlet extraction. In another experiment, silica was treated with TDI-PEG-TDI trimer which was endcapped with methanol, again 2% of PEG was found on silica surface. These suggest that PEG does not react withsilanol group of silica via - -Si- -O- -C ester linkage under current reaction conditions, however, 2% of PEG canbe firmly adsorbed onto silica surface through hydrogen bonding.Quantitative characterization by TGA provides evidence of PEG grafting. Figure 2 shows TGA curves ofilica, TDI-grafted silica, product obtained from the typical procedure and those from the control experimentsmentioned above. As it can be seen from curve 2, TDI moiety decomposes mainly between 150°C and 300°C,while PEG is known to decompose between 300°C and 600°C. Therefore, the amount of PEG in the product canbe calculated based on curve 3, which shows 16.7%, much higher than the maximum adsorption value (2%,curves 4 and 5), revealing the occurrence of PEG grafing through TDI coupling.Results of elemental analyses provide supplementary evidence for PEG grafting. Silica does not contain Cand N. The C and N contents of one of the products are 10.90% and 0.72%, respectively, which correspond to0.32 mmol of TDI grafting and 0.09 mmol of PEG grafting. These values are basically in line with thosecalculated from TGA data, proving that TGA is an appropriate means to quantitate the PEG grafing.Figure 3 shows the TEM microphotographs of the original silica and PEG-grafted silica. There is noobvious change after grafing, though the tendency to agglomerate seems to decrease to some degree, which is incontrast to the case of nan-apatite 10. Silica and nano apatite are polyfunctional, in the presence of a couplingagent such as disocyanate, and crosslinking maybe occur. But in fact, no gelation was observed during thecourse of reaction. The result of crosslinking is much bigger particle size, which is not the case as revealed byTEM observation.Effect of Temperature on Graftingt is well known that temperature may have significant influence on the reactivity of reactants, so wecomprehensively investigated the effect of temperature on grafting in one-step procedure. Three temperatures,20°C, 50°C and 80°C were chosen in our studies. As TDI may pror end hydroxylgroups of PEG for each temperature, we designed three experim中国煤化工in which all thereagents were added simultaneously before the reaction and theC N M H Gtemperature for7 h, (2) TDI pre-grafting procedure, in which silica, TDI, DBTDL, and toluene were first added and sirred for a382B. Wang et al.period of time under the protection of nitrogen, and then PEG was added and reacted for another period of time,(3) TDI- -PEG- -TDI urethane procedure, in which PEG, TDI, DBTDL, and toluene were first added and stirredfor a period of time under the protection of nitrogen, and then silica was added and reacted for another period oftime. In all these experiments, the amounts of the reagents are the same as those mentioned in the experimentalpart. It is worthy to note that the molar ratio ofTDI to PEG is kept at 2:1.100nm50 nmFig. 3TEM microphotographs of original silica (a) and PEG-grafted silica (b)By comparing the results obtained from the three experiments, we can figure out whether one-step procedureis more similar to experiment (2) or to experiment (3). The TGA curves of the products from the threeexperiments at three different temperatures are shown in Figs. 4, 5 and 6.At 20°C, the amount of TDI grafting is 11.6% for TDI pre-grafting procedure (curve 2), and lttle forurethane procedure (curve 3). It is 12.0% for one-step procedure (curve 1), practically the same as that of TDIpre- grafting procedure. This indicates that grafting occurs sequentially just as in TDI pre-grafting procedure, i.e.TDI reacts firstly with silica by using one of its isocyanate groups, and the remaining isocyanate groups providethe sites for PEG grafting (Scheme 1, Process a). In other words, silica is more reactive toward TDI than PEG at20°C. Yosomiga et al." have reported that silanol group on the surface of glass fiber has similar reactivity as theordinary alcohol in terms of reacting with isocyanate. Silanol groups on the silica surface is similar to those onthe surface of glass fiber, and PEG 2000 is less reactive than ordinary alcohol due to its long chain andinsolubility at 20°C. Thus, it seems understandable that silica has a higher reactivity than PEG.1059s-95莓85|葛857575-65200400600 800 1000200 400 -Temperature CCTemperature (C)Fig. 4 TGA curves of the products from one-stepFig. 5 TGA curves of the products from one-procedure (1), TDI pre-grafing procedure (2)中国煤化工汇procedureand urethane procedure (3) at 20°CTYHCNMHGGrafting of PEG onto Nano-silica383At a high temperature (80°C), the amount of TDI grafting is 21.9% for TDI pre -grafting procedure (curve2), which is much higher than that at 20°C. However, the amount of TDI grafting is only 4.5% for one-stepprocedure, which is nearly equal to that obtained from urethane procedure (curve 3), suggesting that graftingoccurs preferentially according to the urethane procedure (Scheme 1, Process b), ie. TDI reacts firstly with PEGto form a trimer (TDI-PEG-TDI) which then reacts with silica to attain the grafting. Note that the molar ratio ofTDIPEG is 2:1.) It is noted that the amount of TDI grafting is, although small, higher than that calculatedaccording to Process b, indicating that only a small amount of TDI reacts with silica while most of TDI reactswith PEG.05-85-75-20o 400600800 1000Temperature (C)Fig. 6 TGA curves of the products from one -step procedure (1), TDI pre~grafing procedure (2) and urethane procedure (3) at 50°C/CH30CN-K_-NcOjCH3( SiO2 ) -OH( SiO2 )0-C-NH-NCOHO(CH2CH2O),HCH3 8SiO2 )-0-C-NH-_》>-NH-C-O(CH.CH2).OHProcess a20CN<_-NCO + HO(CH2CH2O),HoCN-《)-NH-C- -O(CH2CH20)%- C- -NH-》NCO(SiO2)HCHy( SiO2)-o-C-NH--NH- -C- -O(CH2CH2OC -NH《)-NCOProcess bScheme 1. Schematic representation of the reaction mechanism via two possible processesFrom the above discussion, we can conclude that due to the catalytic effect of DBTDL, the reactivities ofPEG and silica are in reverse order as compared with those at 20中国煤化工e mechanism ofDBTDL-catalyzed urethane formation involves both TDI and hy_ition stel12, 131.:MHCNMHG384B. Wang et al.PEG is soluble in the reaction medium at 80°C, and thus much easier to be activated than silica. Therefore, it isunderstandable that DBTDL accelerates the reaction of PEG with TDI than with silica more efficiently.At 50°C, curves 2 and 3 are similar to those at 20°C, suggesting that both TDI- pre-grafting procedure andurethane procedure are not significantly affected by a slight increase of temperature. TDI grafting amounts to9.4% for one-step procedure, 20% less than that of TDI pre-grafting procedure, indicating that grafting occursmainly according to Process a, but Process b is also involved.The one step procedure and the two two- step procedures at the same reaction temperature (Figs. 4, 5& 6) interms of PEG grafting were compared. It is obvious that one-step procedure gives higher amount of PEG grafingthan both of the two two-step procedures. This seems quite reasonable because the systems are moisture-sensitive, and the reactivity of NCO group could be easily destroyed by atmospheric moisture. Using the currentone-step procedure, PEG grafting is typically around 20% and the grafting eficiency for PEG is around 15%,which is much higher than in the case of two-step procedures. For example, in the case of nano-apatite inDMFl0, the grafting efficiency is only about 2%, while PEG grafting is about 8% estimated from the reportedTGA plot. In the case of ferrite in bulk reaction"4, the grafting efficiency is less than 1% while the grafting ofPPG is typically 18.5%. The low grafting efficiency in the reported tw-step procedures were resulted from theuse of large excess amount of polyether diols.Effect of Molecular Weight (MW) of PEG on GrafingTo evaluate the effect of molecular weight of PEG on grafting, a series of reactions using PEG with molecularweight ranging from 400 to 6000 was carried out under the same conditions as described in the typical procedure.The results are listed in Table 1. When the molecular weight of PEG increases from 400 to 6000, the grafting ofPEG increases from 13.3% to 24.8%, however, the number of grafted PEG chains decreases sharply from 0.33mmol/g of silica to 0.04 mmol/g of silica, indicating that the reactivity of the intermediate (TDI-PEG-TDI trimer)decreases considerably with the increasing of chain length of PEG. It is noted that the grafting efficiency of PEGis rather high for low molecular weight PEG (47% for PEG400), and decreases considerably with the increase ofmolecular weight of PEG.Table 1. Effect of MW of PEG on gratingPEG graftingGrating efciency (%)TDI gratingOverall grafingMW/added(g)(%)(mmol/g)(%)400/0.2813.3/0.33478.621.91000/0.715.1/0.15217.222.32000/1.416.70.095.21.86000/4.224.8/0.04CONCLUSIONSPEGs with molecular weights of 400 to 6000 can be successfully grafted onto nano-silica by DBTDL-catalyzeddisocyanate coupling through a one-step procedure. Two grafting processes are proposed. One involves a firstreaction of TDI with silica to give a surface bearing isocyanate groups and a subsequent reaction of thefunctionlized silica with PEG. The other involves a first reaction of TDI with PEG to give a urethane trimer withisocyanate end groups and a subsequent reaction of the functionalized polymer with silica. At low temperature(20°C), grafting reaction proceeds preferably according to the first process. At high temperature (80°C), graftingreaction proceeds preferably according to the second process. At intermediate temperature (50°C), bothprocesses are involved. This reveals that silica is more reactive than PEG at low temperature and less reactivethan PEG at high temperature in DBTDL-catalyzed urethane formation reactions.This work provides a simple and convenient procedure for the synthesis of modified silica having reactiveorganic surface that could potentially be useful in the field of nan中国煤化工MYHCNMHGGrafting of PEG onto Nano silica385REFERENCESHayashi, s.. Takeuchi, Y, Eguchi, M., lida, T and Tsubokawa, N., J. Appl. Polym. Sci, 199, 71 (9): 14912 Tsubokawa, N, Shirai, Y.. Tsuchida, H. and Handa, S., J. Polym. Sci. Part A: Polym. Chem, 1994, 32: 2327Tsubokawa, N. and Ishida, H, Polym. J., 1992, 24: 8094 Werme T. and Patten TE, J. Am. Chem. Soc, 1999, 121:7409Yu, J, Yu, J, Guo, Z.X. and Gao, Y.F, Macromol. Rapid Commun., 2001, 22: 12616 Yu,J.. Yu, J, Gao, Y.F and Guo, ZX., Chinese J. Polym. Sci, 2002, 20: 71Fujiki, K, Sakamoto, M, Sato, T. and Tsubokawa, N, J. Macromol. Sci.~Pure Appl. Chem, 2000, A37 (4): 357Guo, Z.X. and Yu, J, J. Mater. Chem, 2002, 12: 4689 Yoshinaga, K., Bull. Chem. Soc. Jpn, 2002, 75: 234910 Qing, L. and de Wjn Joost, R., Biomaterials, 1989, 19:106711 Yosomiya, R., Morimoto, K. and Suzuki, T., J. Appl. Polym. Sci, 1984, 9: 67112 Kothandaraman, H. and Sultan Nasar, A., J. Macromol. Sci.- Pure Appl. Chem, 1994, A31(3): 33913 Van Der Weij, F.W., J. Polym. Sci. Part A: Polym. Chem, 1981, 19: 38114 Tsubokawa, N. and Maruyama, K.. Colloid Polymer Sci, 1989, 267: 511中国煤化工MYHCNMHG