High-efficiency acetaldehyde removal during solid-state polycondensation of poly(ethylene terephthalate)assisted by supercritical carbon dioxide☆

Zhenhao Xi,Tian Liu,Wei Si,Fenglei Bi,Zhimei Xu*,Ling Zhao

State Key Laboratory of Chemical Engineering,East China University of Science and Technology,Shanghai 200237,China

Keywords:Supercritical CO2 Solid-state polycondensation Poly(ethylene terephthalate)Removal of acetaldehyde

A B S T R A C T The concentration of acetaldehyde(AA)is the main quality index of poly(ethylene terephthalate)(PET)used in food and drink packaging.A new method for AA removal has been developed by using supercritical carbon dioxide(scCO2)during the solid-state polycondensation of PET.The influence factors of AA removal including the temperature,pressure,reaction time and the size of pre-polymer particles are systematically studied in this work.The results indicate that it is a highly efficient way to obtain high molecular weight PET with relative low concentration of AA.Correspondingly,the polymerization degree of PET could increase from 27.9 to 85.6 while the concentration of AA reduces from 0.229×10?6 to 0.055×10?6 under the optimal operation conditions of 230 °C,8 MPa and size of 0.30–0.45 mm.Thermodynamic performance tests show the increasing extent of PET crystallinity due to the fact that the plasticization of scCO2 is not obvious with extended reaction time,therefore the increasing crystallinity has no significant influence on AA removal.SEM observations reveal that the effects of scCO2-induced plasticization and swelling on PET increase significantly with the decrease of prepolymer size,and the surface of PET becomes more loose and porous in favor of the AA removal.

1.Introduction

Poly(ethyleneterephthalate)(PET)is a thermoplastic aromatic polyester with excellent thermodynamic and mechanical properties and has been widely used in fiber industry as well as in food and drink packaging[1].With the rapid development of beverage and beer packaging,the demand for bottle-grade PET increases by double digits per year in recent days.The polymerization degree(Pn)of bottle-grade PET is higher than that of standard fiber products,usually more than 100 in industry.The concentration of acetaldehyde(AA)is the main quality index of PET,especially in bottle-grade PET for food and drink packaging.The residual AA can easily change the taste,color and thermophysical properties of PET,that will adversely affect the quality and application of final products and even cause the environmental and health hazards.Besides,strict regulations have been enacted to establish maximum permissible levels of AA,which is required to be lower than 1×10?6in bottle-grade PET as listed in the national standard GB17931-2003.

For bottle-grade PET with relatively high molecular weight(MW),there are usually two ways to efficiently increase the polymerization degree of PET:the melt polycondensation[2–4]and the solid-state polycondensation(SSP)[5–7].For melt polycondensation,the polymerization degree of polymer is increased by prolonging the reaction time in a specially designed high-viscosity polycondensation device.But the viscosity of bulk melt increases rapidly with the raise of MW resulted in a dramatic grow th of agitation power and great difficulty for devolatilization.The problems mentioned above limit the grow th of Pn,which can only increase to about 100 after melt polycondensation.On the other hand,the long reaction time at relatively high temperature will increase seriously the thermal degradation,which leads to increasing AA concentration up to 100×10?6.For SSP,although it can effectively inhibit the thermal degradation reaction to obtain PET with high Pn and lower concentration of AA by using a SSP tower coupled with a downstream dealdehyder,there are still many problems especially the usage of a lot of normal nitrogen for devolatilization and a long operation time with high energy consumption.

The utilization of supercritical fluids in polymers processing has attracted much attention over the last two decades,such as the crystal transition of syndiotactic polystyrene in supercritical CO2-containing solvent[8–12],foaming of polypropylene[13–16]and PET[17–20]with scCO2,extracting impurities from polymers,etc.[21,22].Currently,environmentally benign fluids such as scCO2and ethyl alcohol(C2H5OH)are attractive alternatives as physical agents due to their unique properties such as non-toxic,quick dissolving,and good solvent characteristics.With relatively low critical temperature and critical pressure(Tc=31.1°C,Pc=7.38 MPa),scCO2in general is one of the most widely used and studied solvents in the extraction process[23,24].During the saturation process,scCO2can quickly diffuse into the free volume between polymer chains and improve the motion ability of long molecule chains to a certain extent.In addition,many volatile impurities with small molecular weight also can be dissolved into scCO2because of the solvent effect.A recent research[25,26]indicated that the SSP of PET was promoted significantly with a periodical scCO2renewing strategy,and the Pn of PET increased from 100 to 150 only at 6 h,while it takes nearly 20 h in the industrial SSP process of PET.Therefore,it can be expected to remove the by-product of AA in PET via SSP process assisted by the periodical scCO2renewing strategy which has not been reported to date.

In this work,the SSP process of PET had been carried out assisted by periodical renewal of scCO2[26].The effects of operation conditions,including the temperature,pressure,reaction time and the size of prepolymer particles,on the molecular weight and the corresponding AA concentration of PET were investigated experimentally by intrinsic viscosity testing and head space gas chromatography,respectively.Correspondingly,the thermodynamic properties and morphology characteristics of PET samples were characterized by differential scanning calorimetry(DSC)and scanning electron micrography(SEM),respectively.This work may provide a new method of SSP without AA removal tower to get high molecular weight PET with low concentration of AA.

2.Experimental

2.1.Materials

Phenol(analytical reagent,purity≥99.0w t%)was purchased from Sinopharm Chemical Reagent Co.Tetrachloroethane(analytical reagent,purity≥99.0w t%)was purchased from Shanghai Lingfeng Chemical Reagent Co.Acetaldehyde(gradient grade,purity 95.5w t%+)and Ethanol(analytical reagent,purity:99.8w t%)were purchased from Shanghai Titan Scientific Co.CO2(purity:99.9w t%)was purchased from Shanghai Chenggong Gases Co,China.H2(purity:99.99w t%),air,N2(purity:99.99w t%)were commercially available from Shanghai Shenzhong Industrial Gases Co.All gases and reagents were used without any further purification.The PET prepolymer(Pn=27.9,[AA]=0.299× 10?6,crystalline=53.59%)used in SSP was provided by Sinopec Shanghai Petrochemical Co.

2.2.Experimental method

As shown in Fig.1,the experimental set-up for the SSP of PET carried out in scCO2was consisted of five elements:the high-pressure autoclave with internal volume of 115 ml,the agitation equipment,the oil bath,the pressurization system and the temperature control system.A two-layer stainless sample cell(5)with 4 cm2square bottom was used to load prepolymers.The pressure of the autoclave was maintained by the pressurization system(2)with a pressure sensor with accuracy of±0.1 MPa.One thermocouple was attached to the temperature control system.

In the SSP process of PET,the system loading with 1.50 g prepolymer was swept by low pressure CO2several times.Subsequently,the autoclave was heated to the reaction temperature in about 5 min and then pressurized to the desired pressure with CO2.After the system reacting 1 h,the pressure was relieved within 2 min,and then the autoclave was re-pressurized to the desired pressure again.The pressurization and depressurization operations,namely periodic renewal of scCO2,were repeated every 1 h[25,26].After the SSP process was completed the system was depressurized to ambient pressure and the autoclave was then cooled down quickly.Finally,the samples were taken out for the subsequent analysis.

The operating conditions of SSP reaction process were carried out as follows:temperature 200–230 °C,pressure 0–10 MPa,pre-polymer size 0.30–3.00 mm,and reaction time 1–14 h.

2.3.Analytical methods

2.3.1.Headspace gas chromatography

The concentration of AA in a PET sample was tested by headspace gas chromatography,and the headspace sampler was Agilent 7697A while GC was Agilent 7890A(with hydrogen flame ionization detector(FID)).The chromatographic column was Agilent HP-PONA Methyl Siloxane 50 m×0.20 mm.The conditions of chromatography were as follow s:N2was used as carrier gas while H2was as combustion gas,the column temperature was 150°C,and the detector temperature was 200°C.The analysis procedure through gradient elevation of temperature was:staying at the constant temperature of 50°C for 2 min and then raising the temperature from 50 °C to 70 °C in 5 min with a flow rate of N2at 2 ml·min?1.

Additionally,the pure AA was also tested by headspace gas chromatography to obtain the standard curves of AA.The concentrations of prepared AA standard solutions and the corresponding standard curves of AAw ere shown in Table S1 and Fig.S1,respectively(see Supplementary Material).

2.3.2.Ubbelohde viscometer

Fig.1.Experimental setup for PET SSP and AA removal.1—gas cylinder;2—pressurization system;3—high-pressure ball valve;4—oil bath;5—high-pressure autoclave;6—temperature sensor;7—pressure sensor;8—temperature control system.

Intrinsic viscosity(IV)of PET sample was measured with an Ubbelohde Viscometer using the mixture of phenol and tetrachloroethane(3/2,w/w)as solvent at 25.0°C.Then the numberaverage molecular weight(Mn)and Pn were calculated as follow s[27]:

2.3.3.Differential scanning calorimetry(DSC)

The thermal behaviors of PET sample were analyzed on a Diamond DSC with the Thermal Analysis Data Station,which provided the thermal analysis data under nitrogen atmosphere.Dried samples of about 5–10 mg placed in aluminum crucible were heated from 50 °C to 280 °C at a heating rate of 40 °C·min?1and held for 5 min to erase any prior thermal history before cooling to 50°C at a cooling rate of 10 °C·min?1.The samples were then reheated up to 280 °Cat a heating rate of 10 °C·min?1.The melting behaviors of PET samples were recorded from the DSC curves.Before the first heating process,the air was replaced by purging the system with N2to avoid the thermal degradation of PET.The crystallinity of PET(Xc)can be calculated by the melting enthalpy ΔHm:

w hereΔHmrepresents the melting enthalpy of sample,ΔHm0represents the melting enthalpy of sample whose crystallinity is 100%.ΔHm0=120 J·g?1[28].

2.3.4.Scanning electron micrography(SEM)

The morphology of PET sample surface was observed on a NOVA Nano SEM 450 scanning electron micrography(SEM)operating at acceleration voltage of 3 kV.All the sample surfaces were made conductive before testing by depositing a layer of gold in 1 min on an Ion Sputter Coater Q150RS,United Kingdom Quorum Technologies.All the scanning electron micrographs were chosen at the same magnification,and the PET samples were ground before testing.

3.Results and Discussion

To verify if the SSP process used in this work has the ability to obtain the bottlegrade PET products,the pre-experiment was firstly conducted at 230°C and 8.0 MPa in scCO2atmosphere,and the corresponding results were shown in Fig.S2(see Supplementary Material).It can be seen that the Pn value of PET particles increases from 27.9 to 149.1 which is similar to that in[26],and the AA concentration also decreases from 0.299×10?6to 0.091×10?6at 14 h.It indicates that the SSP process assisted by scCO2is effective to produce the bottle grade PET.Then,the effects of different operation conditions on Pn and AA concentrations of PET samples using the SSP process assisted by scCO2are investigated and discussed in detail as follows.

3.1.Comparison between scCO2 and normal N2

The corresponding results of AA removal during SSP process of PET in the normal N2and the SSP assisted by scCO2are listed in Table 1.Furthermore,the SSP process in N2atmosphere assisted with vacuuming(200 Pa)was also investigated and the results were shown in Fig.S3(see Supplementary Material).It is clearly shown that the Pn value of PET particles only increases from 27.9 to 35.6 in the normal N2at 4 h,and the AA concentrations decrease from 0.299× 10?6to 0.230× 10?6correspondingly.How ever,the Pn value of PET particles can increase from 27.9 to 63.9 in SSP assisted by scCO2at 4 h,while AA concentrations decrease from 0.299×10?6to 0.124×10?6correspondingly.Fig.S3 also indicates that high Pn of PET with relatively low AA concentration can be preferred to obtain in SSP process assisted by scCO2.That is because of the plasticization and swelling effect of scCO2[23,29,30],which increases the free volume,promotes the diffusion of small volatile molecules in polymer and improves the motion ability of polymer molecular chains and the reactivity between the end-groups.Therefore,the SSP process assisted by scCO2can contribute to obtain high molecular weight PET with low content of AA simultaneously.

Table 1Effect of the normal N2 and the SSP assisted by scCO2 on Pn and the AA concentrations of particles at 230°C and 4 h

3.2.Temperature effect

The effects of temperature on Pn and AA concentration in the SSP process assisted by scCO2were studied at the temperature of 200°C,210 °C,220 °C,and 230 °C,and pressure of 8.0 MPa for the total reaction time of 4 h with periodical renewal of every 1 h,as shown in Fig.2.With the increase of reaction temperature,the Pn value of PET product increases gradually while AA concentration of PET product decreases correspondingly.It is because the motion ability of polymer molecular chains and the reactivity between the end-groups are enhanced with increasing reaction temperature,especially in scCO2atmosphere,which promotes the chain-grow th reaction to obtain high molecular weight PET.Additionally,the side reaction of thermal degradation is also promoted with increasing temperature,which leads to the increase of byproduct concentration.How ever,the high reaction temperature is also in favor of the diffusion of scCO2into the PET matrix,then the dissolution of small volatile molecules(AA,ethylene glycol etc.)into the scCO2can be likewise promoted.As a consequence,those side products can be quickly extracted together through the operation of periodic renewal of scCO2.The scCO2atmosphere can also prevent the thermal degradation reaction of the PET particles which contributes to restrain the formation of AA.Consequently,the SSP of PET assisted by scCO2can efficiently obtain the high molecular weight of PET with a low AA concentration.

Fig.2.Effects of reaction temperature on Pn and AA concentration of PET in SSP(pressure 8 MPa,pressuring and pressure-relief every 1 h,total reaction time 4 h).

3.3.Pressure effect

The effects of pressure on Pn and the concentration of AA during SSP of PET assisted by scCO2were also explored at the pressure of 8 MPa,10 MPa,12 MPa,and temperature of 230°C for the total reaction time of 4 h with periodic renewal of every 1 h,as shown in Fig.3.The Pn and AA concentration of PET can be significantly improved under scCO2atmosphere compared to the raw materials PET,but the Pn of all final PET products are almost the same.Although the plasticization effect of scCO2[26,31]can promote the diffusion of volatile by-products in polymer,the density of PET decreases with the increase of CO2pressure and the diffusion path of by-products increased significantly in the polymer due to the swelling effect of scCO2.Therefore,the contra-concerted effect of plasticization and swelling of scCO2on PET polycondensation process results in the little difference of Pn among the final PET products.

Fig.3.Effects of reaction pressure on Pn and AA concentration of PET in SSP(temperature 230°C,pressuring and pressure-relief every 1 h,total reaction time 4 h).

3.4.Reaction time effect

The effects of reaction time on Pn and the AA concentration of PET in the SSP were studied at the time of 1 h,2 h,3 h,4 h,6 h,10 h,temperature of 230°C and pressure of 8 MPa with periodic renewal of scCO2,as shown in Fig.4.With the increase of reaction time,the Pn of PET product raises gradually while the concentration of AA in PET product decreases.The Pn value of PET is close to 100 after 10 h which already achieves the quality of commercial PET products.Additionally,the concentration of AA in the PET particles always maintains at a very low level(＜0.2×10?6)during the whole process.Compared to the traditional melt polycondensation process of PET with the Pn value equal to 100 and AA concentration close to 100×10?6,it is indicated that the SSP of PET assisted by scCO2can serve as a potential technology to produce PET of low content of AA(＜1× 10?6).

3.5.Pre-polymer size effect

According to Ref.[32],the rate of SSP is much larger with smaller pre-polymer size,therefore it is much easier to produce high MW of PET.The effects of pre-polymer size on the SSP of PET were also studied at the size of less than 0.30 mm(60 mesh),0.30–0.45 mm(40–60 mesh),polymer particle(3.00 mm × 3.00 mm),temperature of 230°C and pressure of 8.0 MPa for 4 h with periodic renewal of scCO2,as shown in Table 2.

Table 2Effects of prepolymer size on Pn and AA concentration of PET in the SSP(temperature 230°C,pressure 8.0 MPa,pressurizing and depressurizing every 1 h)

It is well known that the process of PET polycondensation consisted of four steps:the motion of reactive end-groups,the polycondensation between reactive end-groups,the diffusion of volatile components in polymer melt(internal diffusion),and the diffusion of volatile components from polymer surface to gas phase(external diffusion)[33].Generally,the specific surface area of particle is much greater with a smaller polymer size and the AA removal is much easier.With decreasing size of PET,the internal diffusion path of volatile components in polymer is shortened,meanwhile the increasing specific surface area of particle promotes the external diffusion of AA.As shown in Table 2,the Pn and AA concentration of PET particles are less than those of PET in 0.30–0.45 mm.However,the Pn of PET product is smaller than others when the size of PET is less than 0.30 mm.It is because,as reported,the PET particle with a smaller size is much easier to agglomerate at a higher temperature,which hinders the flow of scCO2in PET matrix,and then inhibits the reaction and AA removal.

3.6.Thermal properties of PET

Fig.4.Effects of reaction time on Pn and AA concentration of PET in SSP(temperature 230°C,pressure 8.0 MPa,pressurizing and depressurizing every 1 h).

Fig.5.DSC curves of PET products.

The thermodynamic properties of PET products at 230°C and 8.0 MPa with different total reaction time of 1 h,2 h,3 h,and 4 h were shown in Fig.5,respectively.The corresponding peaks of the crystallization temperature and the melting temperature,the enthalpies and the degree of crystallization were listed in Table 3.It can be seen that the melting point Tm1and melting enthalpy ΔHm1increase gradually with increasing reaction time,while the melting crystallization temperature Tcand crystalline enthalpy ΔHcinversely decrease,which is consistent with the conclusion of Zheng et al.[34].As mentioned above,the MW of PET gradually increases with increasing reaction time which contributes to the increase of the melting point.Also,the mobility of polymer molecular chains and interior diffusion steps can be strengthened by the plasticization of scCO2with increasing reaction time,which contributes to the reducing of crystallization temperature.Additionally,the crystallinity of PET(Xc1)increases gradually with the increase of reaction time.Compared the crystallinity of Xc1and Xc2,the crystallinities of PET samples with thermal history in scCO2atmosphere are higher than that of eliminated thermal history which is attributed to the plasticization and induced crystallization of scCO2.How ever,the change extent of PET crystallinity is not obvious with increasing reaction time,which means it has no significant influence on AA removing rate.

3.7.Morphology properties of PET

In order to characterize the influence of scCO2on the SSP process under different process conditions,the morphology properties of PET samples were characterized by SEM,and the results were shown in Fig.6.The corresponding operation conditions and results for SEMwere listed in Table 4.With the increase of reaction time,the surface of PET becomes smooth and has a certain extent of orientation which contributes to the increasing crystallinity.Especially,there are some obviously fractured fibers in 3#sample which is because the sample has a good spinnability with the Pn value of PET close to 100.With a lower reaction temperature,the reaction rate of SSP is much slower and the Pn of PET is also decreased,therefore there is no significant difference in the morphology compared to the raw material.The surface of PET generates some small holes gradually,as shown in 5#and 6#PET samples which can be derived from the stronger plasticization and swelling impact of scCO2to PET matrix under higher pressure.Compared the 7#and 8#PET samples,the surface of PET becomes loose and porous while the pore size clearly increases.It indicates that the plasticization and swelling of scCO2to PET matrix are stronger when the size is smaller,which is benefit to by-product removal.How ever,when the size of particle is too small,it is easy to agglomerate under high temperature which causes a negative influence to the SSP process.Therefore,it is necessary to choose the size of PET reasonably for efficient removing of AA through SSP process,and the optimal size is in the range of 0.30–0.45 mm.

Table 3DSC results of PET products

Fig.6.SEM results under different process conditions.

Table 4Process conditions and results corresponding to SEM

In summary,it can be found that only 14 h is needed to reach a Pn of 150 from 27.9 in the SSP process of PET assisted by scCO2,in contrast that it takes nearly 20 h to obtain a Pn of 150 from 100 in the present industrial process using atmospheric hot N2[26].Additionally,the content of AA is always maintained at a lower concentration below 0.30×10?6during the overall SSP process assisted by scCO2,which indicates that there are a few side reactions.How ever,the content of AA needs to be decreased to less than 1×10?6from about 100×10?6in the present industrial process with serious side reactions.It can be concluded that the SSP process assisted by scCO2is a potential technology to obtain the bottle-grade PET products of high MW and low AA concentration.

4.Conclusions

The effects of different operation conditions on the SSP of PET assisted by scCO2,including the temperature,pressure,reaction time and pre-polymer size,are investigated experimentally in detail.By using the SSP assisted by scCO2,the higher molecular weight of PET with lower AA contents is much easier to obtain under the conditions of 230 °C,8.0 MPa and size of 0.30–0.45 mm.The Pn value of PET product gradually increases with increasing reaction time,while the concentration of AA decreases.The melting point,melting enthalpy and crystallinity of PET products increase gradually with the increase of reaction time,while the melting crystallization temperature and crystalline enthalpy decrease,which are attributed to the induced crystallization behavior to PET by scCO2.How ever,the increasing crystallinity has no significant influence on AA removal.Due to the contrasynergistic effect between the plasticization and swelling of scCO2on PET matrix,the pressure has no remarkable influence on the SSP process,while the higher reaction pressure is beneficial to AA removal.Additionally,the plasticization and swelling impact of scCO2to PET particles with a smaller size are stronger.The surface of PET sample becomes loose and porous,which makes AA much easier escape from the PET matrix under the atmosphere of scCO2.Compared to the industrial SSP process for bottle-grade PET with atmospheric hot N2,the scCO2process is a preferable method to obtain the high Pn and low AA content of PET products at a shorter process time.

Supplementary Material

Supplementary data to this article can be found online at https://doi.org/10.1016/j.cjche.2018.03.007.