2 EXPERIMENTAL
2.1 Materials
PBT chips were kindly supplied by Yi Zheng Chemical Fiber Co., China. Dimethyl terephthalate (DMT), ethyl glycol (EG), isophthalate acid (IPA), sebacate acid (SA), zinc acetate (Zn(OAC)₂), and tetrabutyl orthotitanate (TBT) were all reagent grade and used as received.
2.2 Synthesis of PETIS random copolyesters and PBT-PETIS segmented copolyesters  
The synthesis of ternary amorphous random copolyesters can be divided into two steps. The first involves the transesterification of DMT and EG with zinc acetate as a catalyst, and the esterification of SA, IPA and EG, yielding bis(hydroxyethyl) terephthalate (BHET), bis(hydroxyethyl) sebacate (BHES), and bis(hydroxyethyl) isophthalate (BHEI). The second one is the principal reaction leading to the formation of copolyester by means of polycondensation in the presence of tetrabutyl orthotitanate (TBT) as a catalyst.
    The melting transesterification between PBT and PETIS was conducted in a four-neck flask equipped with a mechanically sealed stirrer and condenser under nitrogen atmosphere and vacuum condition less than 133Pa. The melting processing was proceeded isothermally at 255°C for 10min, 20min, 40min, and 60 min, respectively. The process is given in Scheme1.

                                                                    PETIS

                                                                                  PBT-PETIS

2.3 Characterization of PBT-PETIS segmented copolyesters
The intrinsic viscosity was determined at 25±0.1°C in a phenol/1,1,2,2-tetrachloroethane (50/50, w/w) solution with a concentration of 0.5g/100ml by an Ubbelohde viscometer, and the density was measured with a gradient density column in a thermostat at 25±0.1℃ using aqueous solution of sodium bromide. Different Scanning Calorimetry (DSC) measurement was carried out on a Shimadzu-50 apparatus calibrated with Al₂O₃ under nitrogen atmosphere using samples of about 5mg. The sample was heated up to 30°C above the melting temperature at a rate of 10°C/min, and maintained for 5 min to erase the thermal history, and then cooled down at a rate of 10°C/min to room temperature. Wide angle X-ray diffraction measurement (WAXD) was done on melt-pressed films in the reflection mode with a Rigaku diffractmeter using Ni-filtered Cu-Ka radiation. The ¹H-nuclear magnetic resonance spectra (NMR) were recorded with a Varian UNITY-plus400MHz NMR spectrometer by dissolving samples in deuterated chloroform (CDCl₃). Copolyester sample was sandwiched in between two pieces of glass slides followed by melting completely. Then the specimen was rapidly quenched to the predetermined crystallization temperature and maintained for 2 hours. Crystalline morphology of copolyester was investigated using a polarized light microscopy (PLM).

3 RESULTS AND DISCUSSION           
Segmented copolyesters at room temperature appear soft or hard semi-crystalline solids according to compositions. The samples with varying compositions are given in Table 1, where intrinsic viscosity and density are also reported.
     The values of intrinsic viscosity demonstrate that melting transesterification is available and yields copolyesters having molecular weights high enough for usual application. Moreover, because of the high values of intrinsic viscosity, effect caused by the discrepancy of molecular weight on properties can be negligible. The density had a little variation for samples with different reaction time, but increased slightly with increasing PBT content in segmented copolyesters.
    All copolyester samples show lower melting peaks compared to the PBT homopolymer. The depression of melting temperature is due to the limited crystallite size and to the larger amount of crystal imperfection because of the less favourable crystallization conditions^[6]. DSC curves in the heating run reveal double melting peaks for PBT-rich copolyesters, which is however a widespread observation that has been reported in the literatures^[7,8]. Park and Kang^[9] pointed out that the lower one can be attributed to melting of PBT crystal related to ternary random copolyester, and the other to that of pure PBT crystal. Rim and Runt^[10], however, thought that the multiple-melting endotherms may result from reorganization processes. They explained the phenomenon as follows: the low-temperature endotherm reflects the true melting behavior of the as formed crystals while the higher temperature endotherm represents the melting of material that has undergone annealing upon heating. For comparison, endothermic peaks were presented in Fig.1 according to reaction time. Table II collected the summary of corresponding thermal properties. The double melting peaks are marked T_m1 and T_m2 in Table II respectively, and DH_m is the heat of fusion; T_c and DH_c are the maximum and the enthalpy of exothermic peak, respectively.

    From Table 2, we can find that all DSC curves show melting peak whose location depends on both the composition and reaction time. Firstly, melting peaks gradually move to lower temperature region and become wider when the content of PBT segment is increased. High feed fraction of amorphous copolyester PETIS should give rise to relatively long sequence of soft segment, which might favor forming of regular crystals of PBT. On the other hand, relatively short amorphous random copolyester segments, which during melting transesterification are generated from low feed fraction of amorphous copolyester, might act as crystal imperfection. As a result, the melting temperatures are depressed, and this is the reason for wider range of melting peaks. Secondly, for a chosen composition, sample with shorter reaction time revealed higher melting temperature than those of with longer reaction time. Extended reaction time promoted shorter sequence length of polyesters^[10], which result in lower melting temperatures. Moreover, the broadened distribution of sequence length caused by prolonged reaction time is the reason for wide range of melting temperature.
    In the same manner, lower crystallization temperatures are presented in PBT-rich samples as well as samples with longer reaction time. The depression of T_c is seemingly because of the incorporation of short sequence length of noncrystallizable copolyester segments generated by decreased amorphous polyester content and prolonged reaction time.

a ^: not detected

    In the WAXD patterns of PBT, five diffraction peaks, namely, 16.1 ^o, 17.2 ^o, 20.6 ^o, 23.1 ^o, and 25.2 ^o are assigned for the (), (010), (), (100), and () plane, respectively. With the increasing content of amorphous copolyester PETIS, the diffraction peaks tend to be broaden, but still exihibit the PBT-related scattering angles. The patterns for samples with 30% PBT(wt%) at various reaction time shown in Fig.2 confirm that reaction time has intensive influence on melting transesterification processing. The intensity of diffraction peaks become weaker as the reaction time is prolonged, which infers the change from segmented into random by the interchange between two different kinds of polyesters. Attempt has been made to divide patterns into separate Gaussian profiles, and then crystallite dimensions can also be calculated by using Scherrer equation as follows:^[12]
             (1)

Fig.1 melting endotherms of sample with 30% PBT (wt %) (reaction time indicated)
Fig.2 WAXD patterns of sample with 30% PBT (wt %) (reaction time indicated)

    where D is the mean crystallite dimension; q , the Bragg angle; l , the wave length (0.1542nm); b , the half-high breadth at the Bragg scattering angle; and k is a constant (0.89). The calculation was done only for the (010), (), and (100) planes whose Bragg angle is quite clear in the patterns. Table 3 collected the crystallography data. The dimensions of segmented copolyesters are very smaller than those of PBT homopolymer for which the melting behavior provided early indication, but there is a trace discrepancy among segmented copolyester samples.

   
                PETIS

                                        30PBT-40                                                       

    Spherulite morphology of PBT homopolymer has been observed by Stein^[14] to have two types. One leads to growth of unusual spherulite at low temperature, and the other to usual spherulite at high temperature. In copolyesters, the two types of spherulites may be somewhat more indistinct, coexist together, and a transitional spherulite is favoured.
     Early work by Keith and Padden^[15] proposed a parameter to characterize the internal morphology of a spherulite:
          (2)
  where B is the diffusion coefficient for noncrystallizable content; and G, the radial growth rate of the spherulite. For high concentration of noncrystallizable polyester PETIS, since the PETIS repeating units enter the PBT main chain to a minor degree on the process of melting transesterification for causing the long sequence length of PETIS segments and the high growth rate of crystals, may it be supposed that, in the formation of the crystals, the spherulites not only pack themselves densely, but also form finer fibrils. Consequently, the value of s decreases and optical micrograph of PBT-PETIS samples reported in Fig.4 showed nice spherulite images in copolyesters with 20%PBT and 30%PBT. Conversely, the spherulitic character is not evident for samples with relatively lower concentration of noncrystallizable polyester.

[ Back ] [ Home ] [ Up ] [ Next ] Mirror Site in  USA  Europe  China  ChinaNet   GBNet