Acetal Copolymer from the Copolymerization of Trioxane and Ethylene

Ind. Eng. Chem. Res. 2000, 39, 2275-2280

2275

Acetal Copolymer from the Copolymerization of Trioxane and Ethylene Oxide Hajime Nagahara, Kenji Kagawa, Katsuhiko Hamanaka, Kohichi Yoshida, Toshiyuki Iwaisako, and Junzo Masamoto*,† Technical Research Laboratory, Asahi Chemical Industry Company, Ltd., Samejima, Fuji 416-8501, Japan

The copolymerization of trioxane and ethylene oxide in the bulk polymerization state using highly purified trioxane as the starting material, boron trifluoride dibutyl ether as the initiator, and methylal as the chain-transfer agent was investigated for the industrialization of the acetal copolymer, and the various important polymerization conditions were checked. The acetal exchange reaction during the copolymerization gave a uniform acetal copolymer with randomized comonomer distribution. To obtain the polymer with a high degree of stability, several checks were made. The important points were the high purification of the monomer and complete deactivation of the catalyst residue. Thus, the polymer as polymerized showed a degree of stability over 99.5%. Introduction Acetal resin is a term used to describe high molecular weight polymers and the copolymers of formaldehyde. First commercialized as a homopolymer in 1960 by Du Pont, acetal resins are engineering thermoplastics which have found broad use in traditional metal applications.1-3 Shortly thereafter, Celanese researchers developed an acetal resin based on the copolymerization of trioxane and cyclic ethers, such as ethylene oxide.4 In 1962, a commercial plant began producing this acetal copolymer. Since then, the rapid expansion of acetal resin production has occurred worldwide. Up to 1971, Du Pont, Celanese, and Celanese joint ventures have been the sole producers of acetal resins. In 1972, Asahi Chemical started to produce the acetal homopolymer utilizing the world’s third type of polyacetal technology.5 Asahi Chemical also industrialized the acetal copolymer in 1985. Asahi Chemical also industrialized the acetal block copolymer as the third type poly(oxymethylene).6 At present, the annual demand of polyacetal resin in the world is about 400 000 ton. As for our commercialization of the acetal homopolymer, we have already reported the polymerization of formaldehyde using tetravalent organo-tin compounds as catalysts, which were more active than conventional catalysts such as quaternary ammonium salt, some of which gave a polymer with the narrow molecular weight distribution of Mw/Mn ) 2 and provided excellent mechanical properties.7 We also reported the initiation mechanism of the copolymerization of trioxane and ethylene oxide, which was newly found in our research.8 This paper is concerned with our research and development for the industrialization of the acetal copolymer by the copolymerization of trioxane and ethylene oxide. * To whom correspondence should be addressed. † Department of Polymer Science and Engineering, Kyoto Institute of Technology, Matsugaski, Sakyo-ku, Kyoto 6068585, Japan. Present address: Department of Management Science, Fukui University of Technology, Gakuen, Fukui-shi 910-8505, Japan. E-mail: [email protected].

It is well-known that polyacetals have an unstable end group. For example, the as-polymerized acetal homopolymer has thermally unstable hydroxyl end groups, thus the end-capping process of acetylation using acetyl anhydride is necessary. It was also the same situation for the acetal copolymer. In the acetal copolymer, the polymerized polymer has a thermally unstable fraction, -CH2CH2O(CH2O)nH of nearly 10 wt %,9 and this thermally unstable fraction -(CH2O)nH was unzipped to a stable oxyethylene unit, -CH2CH2OH, following the stabilizing section by the melt hydrolysis process. However, this process not only produced the loss of polymer but also provided a complicated stabilizing process. This end-group stabilizing process is the prevalent common knowledge for the acetal copolymer in polymer chemistry and the chemical industry and is documented in the textbook.10-14 For the acetal homopolymer, we, at Asahi Chemical, already succeeded in the first industrialization of endcapping during the polymerization of formaldehyde using highly purified formaldehyde and acetic anhydride as the chain-transfer agent or end-capping agent (eq 1).6

Thus, the target of our research and development of the acetal copolymer was to polymeirize the polymer with a high degree of stability. Our concept to obtain the highly stabilized polymer was based on the end capping during the polymerization. Here we tried to obtain the polymer polymerized for the acetal copolymer with no substantial unstable fraction. If this trial succeeds, it may change the conventional common sense of polymer science and the chemical industry for the acetal copolymer. We then tried the complete end capping during the copolymerization of trioxane and ethylene oxide using methylal as the chain-transfer agent or end-capping agent. This was the world’s first trial for the acetal copolymer. Our concept is as follows: If the feed monomer contains some impurities such as water, the end group of the polymer may be the thermally unstable hydroxyl group (eq 2). (In eq 2,

10.1021/ie990826d CCC: $19.00 © 2000 American Chemical Society Published on Web 05/31/2000

2276

Ind. Eng. Chem. Res., Vol. 39, No. 7, 2000

ethylene oxide as the comonomer was omitted for simplicity.) If the monomer is highly purified and using methylal as the chain-transfer agent, the end group of the obtained polymer may be the thermally stable methoxy group (eq 3). (In eq 3, ethylene oxide as the comonomer

was omitted for simplicity.) Our research and development was the realization of this new concept. This might be the first trial in the world. Experimental Section Material. Trioxane was synthesized by the trimerization of formaldehyde in the aqueous state, distilled, and then extracted with benzene. The new method for the synthesis of trioxane will be reported soon.15 The extracted trioxane was purified by distillation according to the method described in our patents.16,17 The water content in the purified trioxane was below 1 ppm. Methylal, which was used as the chain-transfer agent, was dehydrated by using 4A molecular sieves. Boron trifluoride dibutyl ether was distilled under reduced pressure, sealed in a glass ampule, and then stored in a refrigerator. It was opened before use and diluted 50 times with cyclohexane when it was used. Ethylene oxide in a 5-kg bomb was used without further purification. Polymerization. Polymerization was done in the bulk state. For the polymerization on a small scale, we used a 100-mL glass ampule. We sometimes used a 2-L table kneader with two σ-shaped mixers for the batch process experiment. We also sometimes used a 5-in. (diameter, 12.5 cm; length/diameter, 7) kneader with lens-shaped paddles for the continuous polymerization experiment. The continuous kneader was made by Kurimoto Tekko, and its tradename was the KRCKneader. Into molten trioxane, methylal was mixed and gaseous ethylene was absorbed, and into this mixture, the cyclohexane solution of BF3‚OBu2 was injected. Deactivation of the Catalyst. For the case of ampule polymerization, the obtained polymer was pulverized in a mortar and immersed in hexane with tributylamine. For the table kneader and the 2- and 5-in. kneader polymerizations, the obtained polymer was immersed in water containing 1% triethylamine. After the deactivation of the catalyst residue, the polymer was vacuum-dried or dried in a nitrogen atmosphere. For the 5-in. kneader polymerization, the as-polymerized polymer was crushed using an in-line mixer and deactivated in 1 wt % aqueous triethylamine for 120 min. The deactivated polymer powders were dried in a paddle drier at 120 °C under a nitrogen atmosphere. Measurement. The reduced viscosity was measured in a 1,1,2,2-tetrachloroethane-p-chlorophenol mixed solvent (1/1 weight ratio) at 60 °C.

Figure 1. Polymer yield versus polymerization time. Initial ethylene oxide concentration: 4.5 × 10-2 mol/mol of trioxane. Methylal feed: 4.5 × 10-3 mol/mol of trioxane. BF3‚OBu2: 7 × 10-5 mol/mol of trioxane. Polymerization temperature: 70 °C. Polymerization in a glass ampule.

The stability degree of the polymerized polymer was calculated based on the residue of the polymer (wt %) at 222 °C after 50 min of heating in a vacuum. The 1H NMR measurements were done using 270MHz NMR (JEOL GX-270 FT-NMR) and 500-MHz NMR (Bruker ARX-500) spectrometers. The polymer was dissolved in 1,1,1,3,3,3-hexafluoro-2-propanol-d2 and measured at 50 °C. Results and Discussion We first checked the copolymerization behavior of trioxane and ethylene oxide. Some interesting phenomena were observed during the copolymerization. These phenomena were explained by the acetal exchange reaction. Polymerization behavior versus time is shown in Figures 1-3. Figure 1 shows the polymer yield versus the polymerization time. The polymer yield first increased rapidly with time and then gradually increased. The polymer yield for a 30-min polymerization was nearly 90%. Figure 2 shows the change in the reduced viscosity versus time. The reduced viscosity decreased with time and finally reached a constant value. The abrupt decrease in the reduced viscosity with time and reach in a constant value were thought to be due to the acetal exchange reaction, as shown in eq 4, which randomized the molecular length.

In Figure 2 the change in the number-average molecular weight, which was determined by the end-group measurement, is also shown. The number-average molecular weight was almost constant, though the reduced viscosity of the polymer decreased with time. These phenomena strongly suggested the acetal exchange reaction. The change in the degree of stability of the polymer with time is shown in Figure 3. The degree of polymer

Ind. Eng. Chem. Res., Vol. 39, No. 7, 2000 2277

Figure 4. Change in the comonomer content versus polymerization yield. Initial ethylene oxide concentration: 4.5 × 10-2 mol/ mol of trioxane. BF3‚OBu2: 7 × 10-5 mol/mol of trioxane. Polymerization temperature: 70 °C. Polymerization in a glass ampule.

Figure 2. Change in the reduced viscosity and number-average molecular weight versus polymerization time. Polymerization conditions are the same as those in Figure 1. The number-average molecular weight was calculated from the end group of the methoxy group analysis by the 1H NMR method (see Figure 8).

Figure 3. Change in the degree of polymer stability and melting point versus polymerization time. Polymerization conditions are the same as those in Figure 1.

stability increased with the polymerization time and finally reached a constant value. This phenomenon suggested that, in the early stage of copolymerization, the acetal exchange reaction was inadequate; thus, the comonomer distribution was not randomized and produced a lower degree of stability.

In Figure 3 the change in the melting point of the polymer with polymerization time is also shown. The melting point was determined by the second scan of differential scanning calorimetry (DSC) measurement. The melting point increased with time and finally reached a constant value. These phenomena come from the fact that the comonomers trioxepane and dioxolane, which come from 1,3,5,7-tetraoxacyclononane, the direct reaction product of trioxane and ethylene oxide,8 have higher reactivity than trioxane, and these comonomers polymerized earlier than trioxane. Thus, the polymer obtained in the early stage showed a lower melting point, causing the melting point to change with time. To show this phenomenon more clearly, the change in the comonomer (ethylene oxide) content with the polymerization yield was plotted in Figure 4. During the early stage of the copolymerization, the obtained polymer consisted of almost only the comonomer, and with the polymerization yield, the comonomer content decreased and finally reached a constant value. The decrease in the reduced viscosity to a constant value, an almost constant value of the number-average molecular weight during the polymerization, an increase in the polymer stabilty with time, and the early consumption of the comonomer are characteristic phenomena of the copolymerization of trioxane and ethylene oxide for the acetal exchange reaction during the copolymerization. Figure 5 shows the effect of water content in the monomer on the stability of the obtained polymer. With an increase in water, the thermal stability of the polymer clearly decreased. Thus, to get highly stabilized poly(oxymethylene), it was necessary to use highly purified trioxane. Figure 6 shows the effect of the ethylene oxide concentration on the stability of the polymer. When the ethylene oxide decreased, the stability of the polymer decreased. The direct reaction between trioxane and ethylene oxide to form 1,3,5,7-tetraoxacyclononane and 1,3,5,7,10-pentaoxacyclododecane has already been discussed.8 Thus, with a decrease in the ethylene oxide concentration, a decrease in the induction period, the time needed for the abrupt polymerization start, was observed. On a commercial basis, polymers with various molecular weights are needed. Thus, we checked the molecular weight control by the methylal feed. Figure 7 shows the molecular weight control using methyal as

2278

Ind. Eng. Chem. Res., Vol. 39, No. 7, 2000

Figure 5. Effect of the water content on the degree of polymer stability. Initial ethylene oxide concentration: 4.5 × 10-2 mol/mol of trioxane. Methylal feed: 5 × 10-3 mol/mol of trioxane. BF3‚OBu2: 7 × 10-5 mol/mol of trioxane. Polymerization temperature: 70 °C. Polymerization in a glass ampule.

Figure 6. Effect of the ethylene oxide content on the degree of polymer stability. Methylal feed: 5 × 10-4 mol/mol of trioxane. BF3‚OBu2: 7 × 10-5 mol/mol of trioxane. Polymerization temperature: 70 °C. Polymerization time: 30 min. Polymerization in a glass ampule. TO: trioxane.

Figure 7. Effect of the methylal feed on the reduced viscosity of the obtained polymer. Initial ethylene oxide concentration: 4.5 × 10-2 mol/mol of trioxane. BF3‚OBu2: 7 × 10-5 mol/mol of trioxane. Polymerization temperature: 70 °C. TO: trioxane.

the chain-transfer agent. The reduced viscosity of the polymer decreased with an increase in the methylal feed. It was obvious that methylal acted as the chaintransfer agent. In Figure 8, a typical 1H NMR spectrum of the acetal copolymer using methylal as the chain-transfer agent or the end-capping agent is shown. A clear signal of the methoxy end group, which came from the chain-transfer reaction of methylal, was found at 3.50 ppm. An

Figure 8. Typical 1H NMR spectrum of the copolymer from trioxane and ethylene oxide obtained using methylal as the chaintransfer agent. Ethylene oxide concentration: 4.5 mol % trioxane. Methylal: 3 × 10-3 mol/mol of trioxane. BF3‚OBu2: 7 × 10-5 mol/ mol of trioxane. Polymerization temperature: 70 °C. Fairly good agreement was observed between the methoxy group found (2.5 × 10-3 mol/-OCH2 unit) and the methoxy group calculated (2.3 × 10-3 mol/-OCH2 unit). M denotes oxymethylene, and E denotes oxyethylene.

Figure 9. Effect of the methylal feed on the methoxy group found in the polymer. Polymerization conditions are the same as those in Figure 9. TO: trioxane.

oxymethylene unit which connects the oxymethylene linkage and methoxy end group is observed at 4.82 ppm. The oxymethylene main-chain linkage is observed at 5.00 ppm. In Figure 9, the effect of the methylal feed on the methoxy group found in the polymer was plotted. The ratio of the methoxy signal (3.50 ppm) to the oxymethylene signal (5.00 ppm) showed that almost all of the methylal fed to the polymerization system became attached to the polymer end group. Thus, both polymers’ end groups were capped by the methoxy group. However, though using highly pure trioxane, the maximum degree of stability was limited to around 99.0%. However, the value of 99.0% was indeed a high value, considering the 10 wt % unstable fraction data reported by the Hoecst group.9 The effect of the methylal feed on the degree of polymer stability was investigated. The degree of polymer stability was almost constant around 98.5% with the addition of methylal. This fact showed that methylal was an excellent chain-transfer agent or end-capping agent. For the commercial production of the acetal copolymer, the increase in polymer yield was an important factor. Thus, we tried to obtain a polymer with a high polymer yield by extending the polymerization time. However, we observed a curious phenomenon such that the degree of stability of the polymer began to decrease

Ind. Eng. Chem. Res., Vol. 39, No. 7, 2000 2279

Figure 10. Effect of the polymer yield on the degree of polymer stability: 9, polymerization in a glass ampule; b, polymerization in a 2-L table kneader. Ethylene oxide concentration: 4.5 mol % trioxane. BF3‚OBu2: (0.7-1.4) × 10-4 mol/mol of trioxane. Polymerization temperature: 70 °C.

Figure 11. Effect of the feed catalyst concentration on the degree of polymer stability. Initial ethylene oxide concentration: 4.5 × 10-2 mol/mol of trioxane. Methylal feed: 5 × 10-4 mol/mol of trioxane. Polymerization temperature: 70 °C. Polymerization time: 30 min. Polymerization in a glass ampule. Polymer yield: from 96% to 100%. TO: trioxane.

when the polymer yield exceeded 90%. In Figure 10, the effect of the polymer yield on the degree of polymer stability was plotted. At first, the degree of stability increased with an increase in the polymer yield, and then the degree of stability showed a maximum value at the polymer yield from 85 to 92% and gradually decreased with increasing polymer yield. In the case of low polymer yield, an insufficient acetal exchange reaction might be postulated, assuming that the comonomer distribution might not be random, and thus the degree of polymer stability might be low. However, the increase in the polymer yield that decreased the degree of stability was thought to be very unusual. We then assumed that deactivation of the catalyst might be insufficient. In the case of the high polymerization yield, permeation of the base to acid catalyst residue might be difficult; thus, some of the catalyst residue might be activated and decompose the polymer. We further checked the effect of the feed catalyst concentration on the degree of polymer stabilty. Figure 11 shows the effects of catalyst concentration on the stability degree. The polymer yield was almost 100% if the catalyst concentration was over 1 × 10-4 mol/mol of trioxane. The polymer stability degree decreased with

Figure 12. Effect of the polymer particle size on the degree of polymer stability: 9, deactivated at 40 °C; b, deactivated at 80 °C. Ethylene oxide concentration: 4.5 mol % trioxane. BF3‚OBu2: 0.7 × 10-4 mol/mol of trioxane. Polymerization temperature: 70 °C. The initial polymer was obtained during continuous polymerization using a 5-in. KRC-Kneader. Polymerized polymer was crushed using an in-line mixer and deactivated in 1 wt % aqueous triethylamine for 120 min. Deactivated polymer powders were dried in a paddle drier at 120 °C under a nitrogen atmosphere: 100 mesh, 148 µm; 60 mesh, 250 µm; 40 mesh, 420 µm; 20 mesh, 840 µm; 12 mesh, 1400 µm.

an increase in the catalyst concentration. This fact suggested that the catalyst residue affected the polymer stability degree. To clarify this point more clearly, we tried to complete the deactivation of the catalyst residue. Figure 12 shows the effect of the polymer’s particle size on the degree of polymer stability. In this experiment, the polymer obtained using a 5-in. continuous kneader was used. With a decrease in the polymer particle size, the degree of stability increased. It was also observed that the deactivation temperature affected the stability degree of the polymer. With an increase in the deactivation temperature, an increase in the degree of stability was observed. Thus, we found an effective deactivation of the catalyst residue by decreasing the polymer particle size and increasing the deactivation temperature. We first thought the degree of polymer stability should depend on the concentration of the catalyst residue; that is, the smaller the particle size, the lesser the concentration of the catalyst residue. We then checked the effect of the polymer particle size on the concentration of the fluorine anion as the catalyst residue in the polymer; however, a small polymer particle size and a large polymer particle size showed the same fluorine anion concentration of around 10 ppm. Thus, we found that the concentration of the fluorine anion as the catalyst residue was quite independent of the polymer particle size. The completely deactivated polymer showed a thermally unstable fraction below 0.5%. We also confirmed that, in the case of a polymer yield over 95%, if the polymer particles were finely crushed, complete deactivation occurred and a high degree of polymer stability was attained. Thus, the deactivation of the catalyst residue was the main factor for obtaining a high degree of stability if the trioxane monomer was highly purified. The completely deactivated polymer showed a thermally unstable fraction below 0.5%. This value might be remarkable, considering the fact that the conventional aspolymerized acetal copolymer usually contained a 10 wt % thermally unstable fraction.9 As a conclusion, complete end capping during the copolymerization of trioxane in the presence of methylal

2280

Ind. Eng. Chem. Res., Vol. 39, No. 7, 2000

was accomplished by using highly purified trioxane and complete deactivation of the catalyst residue. This new invention provides us an as-polymerized polymer with no substantial unstable fraction and gives us a simple polymer manufacturing process and high-quality product. Complete end capping during the polymerization will change the common knowledge about the acetal copolymer (trioxane-ethylene oxide copolymer) currently written in textbooks. Conclusions For the industrialization of the acetal copolymer from trioxane and ethylene oxide, a complete end-capped polymer during the polymerization was targeted. The following conclusions were obtained: 1. During the polymerization, an acetal exchange occurred, and the comonomer distribution was randomized. 2. To get a polymer with a high degree of stability, the water content in monomer should be minimized. 3. Methylal was a useful chain-transfer agent to control the molecular weight of the polymer, and both end groups were capped with the methoxy group by methylal during the polymerization. 4. Deactivation of the catalyst residue was an important factor to get a polymer with a high degree of stability. One of the best ways for the deactivation was the small polymer particle size. Elevated temperature was also useful for catalyst deactivation. 5. An almost completely end-capped polymer (the degree of stability was over 99.5%) was attained. Literature Cited (1) Schweitzer, C. E.; MacDonald, R. N.; Punderson, J. O. Thermally Stable High Molecular Weight Polyoxymethylenes. J. Appl. Polym. Sci. 1959, 1, 158. (2) Koch, T. A.; Lindvig, P. E. Molecular Structure of High Molecular Weight Acetal Resins. J. Appl. Polym. Sci. 1959, 1, 164. (3) Linton, W. H.; Goodman, H. H. Physical Properties of High Molecular Weight Acetal Resins. J. Appl. Polym. Sci. 1959, 1, 179.

(4) Walling, C.; Brown, F.; Bartz, K. (Celanese). Copolymers. U.S. Patent 3,027,352, 1962. (5) Kobayashi, Y.; Suzuki, I.; Ishida, S. How Asahi Makes Polyacetals. Hydrocarbon Process. 1972, 51 (11), 111. (6) Matsuzaki, K.; Hata, T.; Sone, T.; Masamoto, J. New Polyacetal Process for the Polymerization of Formaldehyde in the Presence of a Chain Transfer Agent. Bull. Chem. Soc. Jpn. 1994, 67, 2560. (7) Matsuzaki, K.; Masamoto, J. Polymerization of Formaldehyde Using Tetravalent Organo-Tin Compounds as an Initiator. Ind. Eng. Chem. Res. 1998, 37, 1729. (8) Nagahara, H.; Kagawa, K.; Iwaisako, T.; Masamoto, J. Initiation Mechanism of the Copolymerization of 1,3,5-Trioxane and Ethylene Oxide. Ind. Eng. Chem. Res. 1995, 34, 2515. (9) Burg, K.; Schlaf, H.; Cherdon, H. Influence of Various Initiators on Homo- and Copolymerization of Trioxane. Makromol. Chem. 1971, 145, 247. (10) Barker, S. G.; Price, M. B. Polyacetals; Butterworth & Co. Ltd.: London, 1970. (11) Persak, K. J.; Blair, L. M. Acetal resins. In Kirk-Othmer Encyclopedia of Chemical Technology, 3rd ed.; Martin, G., David, E., Eds.; Wiley: New York, 1978. (12) Dolce, T. J.; Grates, J. A. Acetal Resins. In Encyclopedia of Polymer Science and Engineering; Mark, H. F., Bikales, N. M., Overberger, C. G., Menges, G., Kroschwitz, J. I., Eds.; John Wiley & Sons: New York, 1985. (13) Sextro, G. Polyoxymethylenes. In Ulmann’s Encyclopedia of Industrial Chemistry; VCH Publishers: 1992. (14) Vairon, J.-P.; Spassky, N. Industrial Cationic Polymerizations. In Cationic Polymerization; Matyjaszewski, K., Eds.; Marcel Dekker: New York, 1996. (15) Masamoto, J.; Hamanaka, K.; Nagahara, H.; Yoshida, K.; Kagawa, K.; Iwaisako, T.; Komaki, H. New Process for 1,3,5Trioxane Synthesis from Aqueous Formaldehyde Solution Using Heteropolyacids as a Catalyst. Ind. Eng. Chem. Res., to be submitted. (16) Hamanaka, K.; Iwaisako, T.; Masamoto, J.; Yoshida, K. (Asahi Chemical). Process for Separating of Trioxane. U.S. Patent 4,332,644, 1982. (17) Masamoto, J. (Asahi Chemical). Process for the Purification of Trioxane. WO 98/13362, 1998.

Received for review November 15, 1999 Revised manuscript received February 28, 2000 Accepted February 29, 2000 IE990826D