Plasticization and cocrystallization in LLDPE/wax blends

The structure and thermal properties of linear low‐density polyethylene (LLDPE)/medium soft paraffin wax blends, prepared by melt mixing, were investigated by differential scanning calorimetry (DSC) and small‐ and wide‐angle X‐ray scattering (SAXS and WAXS). The blends form a single phase in the melt as determined by SAXS. Upon cooling from the melt, two crystalline phases develop for blends with more than 10 wt % wax characterized by widely different melting points. The wax acts as an effective plasticizer for LLDPE, decreasing both its crystallization and melting temperature. The higher melting point crystalline phase is formed by less branched LLDPE fractions. On the other hand, the lower melting point crystalline phase is a wax‐rich phase constituted by cocrystals of extended chain wax and short linear sequences of highly branched LLDPE chains. The presence of cocrystals was evidenced by standard DSC results, successive self‐nucleation and annealing (SSA) thermal fractionation and by the detection of a new SAXS signal attributed to the lamellar long period of the cocrystals. © 2016 Wiley Periodicals, Inc. J. Polym. Sci., Part B: Polym. Phys. 2016 , 54, 1469–1482     

Phase change materials based on low-density polyethylene/paraffin wax blends

Phase change materials, based on low-density polyethylene blended with soft and hard paraffin waxes respectively, were studied in this paper. DSC, DMA, TGA and SEM were employed to determine the structure and properties of the blends. The blends were able to absorb large amounts of heat energy due to melting of paraffin wax, whereas the LDPE matrix kept the material in a compact shape on macroscopic level. The hard paraffin wax was, however, much more miscible with LDPE because of co-crystallization than the soft paraffin wax. LDPE blended with hard paraffin wax degrades in just one step, while blends containing soft paraffin wax degrade in two distinguishable steps. SEM showed completely different morphology for the two paraffin waxes and confirmed the lower miscibility of LDPE and soft paraffin wax. DMA analyses demonstrated the toughening effect of the waxes on the polymer matrix. This technique was also used to follow the thermal expansion as well as the dimensional stability of the samples during thermal cycling. The most visible expansion could be seen in the first cycle, probably due to a totally different thermal history of the sample. With further cycling the dimensions stabilized after two and four cycles for soft and hard paraffin wax, respectively. Controlled force ramp testing on DMA confirmed poor material strength of the blends containing soft wax, especially at temperatures above wax melting.     

Poly(ethylene-co-vinyl alcohol) and poly(methyl methacrylate) blends: Phase behavior and morphology

In this work blends of poly(ethylene-co-vinyl alcohol) (EVOH) with different ethylene contents (27, 32, 38 and 44 mol%) and poly(methyl methacrylate) (PMMA) were prepared by mechanical mixing in the melted state. The miscibility and melting behavior as a function of blend composition and the ethylene content in EVOH copolymers were investigated by means of differential scanning calorimetry (DSC) and dynamic mechanical thermal analysis (DMTA). The morphology of the cryofractured surfaces was examined by scanning electron microscopy (SEM). DSC and DMTA data show that EVOH/PMMA blends are immiscible, independent of EVOH and blend composition. The SEM analysis in agreement with DMTA analysis indicates that the morphology of phases depends on the blend composition, with phase inversion occurring as the concentration of one or other polymer component increases. However, the copolymer composition apparently does not affect the domain size distribution for blends containing 20 wt% of EVOH or 20 wt% of PMMA. A better phase adhesion is observed mainly for blends with 50 wt% of each polymer component.     

Raman,X-ray diffraction and differential scanning calorimetry studies of the melt-induced changes in uncompatibilized and compatibilized blends of high-density polyethylene and nylon 12

The crystallinity of non-molten and pre-molten uncompatibilized and compatibilized polymer blends of high density polyethylene (HDPE)/Nylon 12 have been investigated by using FT-Raman spectroscopy, differential scanning calorimetry (DSC), and wide angle X-ray diffraction (WAXD). The FT-Raman, DSC, and WAXD measurements have revealed that the crystallinity of HDPE of both uncompatibilized and compatibilized blends increases upon melting except for the compatibilized blend with the Nylon 12 content of 80 wt%. The degree of the increase is significantly larger for the uncompatibilized blends than the compatibilized blends. The FT-Raman data suggests that the crystallinity of the compatibilized blend with the Nylon 12 content of 80 wt% decreases slightly after melting. It is very likely that the compatibilized polymer blends are well oriented during the melting and molding process by an extruder because of the existence of maleic anhydride (MAH)-grafted copolymer as a compatibilizer. In contrast, it seems that the uncompatibilized polymer blends are not so much oriented and have more amorphous phase in the Nylon 12 rich region before melting. Thus, the melting process induces a recrystallization process.     

Applications of some thermo-analytical techniques to glasses and polymers

Thermal characterization of materials provides conclusions regarding the identification of materials as well as their purity and composition, polymorphism, and structural changes. Analytical experimental techniques for thermal characterization comprise of a group of techniques, in which physical properties of materials are ascertained through controlled temperature program. Among these techniques, traditional differential scanning calorimetry (DSC) is a well-accepted technique for analyzing thermal transitions in condensed systems. Modulated DSC (MDSC) is used to study the same material properties as conventional DSC including: transition temperatures, melting and crystallization, and heat capacity. Further, MDSC also provides unique feature of increased resolution and increased sensitivity in the same measurement. “Hot disk thermal constant analyzer”, based on Transient Plane Source (TPS) technique, offers simultaneous measurement of thermal transport properties of specimen, which are directly related to heat conduction such as thermal conductivity (λ) and thermal diffusivity (χ). This method enables the thermal analysis on large number of materials from building materials to materials with high thermal conductivity like iron. The temperature range covered so far extends from the liquid nitrogen point to 1000 K and should be possible to extend further. This review also presents some interesting results of phase transition temperature of miscible (CPI/TPI) and immiscible (PS/PMMA) polymeric systems carried out through dynamic mechanical analyzer along with the thermal transport properties obtained for cis-polyisoprene (CPI), trans-polyisoprene (TPI), and their blends determined by TPS technique.     

Isomorphic behavior of poly(aryl ether ketone) blends

Blends of various poly(aryl ether ketones) have been found to exhibit a range of miscibility and isomorphic behavior. This range is dependent on molecular weight; however, for poly(aryl ether ketones) with number-average molecular weight of 20,000, this range is about ±25% difference in ketone content. All miscible blends exhibit isomorphism, and all immiscible blends exhibit no evidence of isomorphism. The dependence of the glass transition temperature T_g versus composition exhibits a minimum deviation from linearity whereas the melting temperature T_m versus composition exhibits a pronounced maximum deviation from linear behavior. The crystalline melting point versus composition for isomorphic blends is considerably different than for random copolymers with isomorphic units. Homopolymers and random copolymers exhibit a melting point that is a linear function of ketone content (increasing ketone content increases T_m). For blends, the melting point is essentially the same as that of the higher melting constituent until high levels of the lower melting constituent are present. The observed melting point versus composition behavior will be interpreted using classical theory to calculate the components of the liquid and crystalline phase compositions. As a miscible blend is cooled from the melt, essentially pure component of the highest melting point crystallizes out of solution, as predicted by calculated solid-liquid phase diagrams. This occurs until the crystallization is complete owing to spherulitic impingement. At high concentrations of the lower melting constituent, lower melting points will be observed because the highest melting constituent will be depleted before the crystallization is complete. In many miscible blends involving addition of an amorphous polymer to a crystalline polymer, the degree of crystallinity of the crystalline polymer has been shown to increase. On the basis of evidence presented here, it is hypothesized that dilution by a miscible, amorphous polymer allows for a higher level of crystallinity.     

Multivariable structural characterization of semicrystalline polymer blends by small‐angle light scattering

A new multi‐variable‐measurement approach for characterizing and correlating the nanoscale and microscale morphology of crystal‐amorphous polymer blends with melt‐phase behavior is described. A vertical small‐angle light scattering (SALS) instrument optimized for examining the scattering and light transmitted from structures ranging from 0.5 to 50 μm, thereby spanning the size range characteristic of the initial‐to‐late stages of thermal‐phase transitions (e.g., melt‐phase separation and crystallization) in crystal‐amorphous polymer blends, was constructed. The SALS instrument was interfaced with differential scanning calorimetry (DSC), and simultaneous SALS/DSC/transmission measurements were performed. We show that the measurement of transmitted light and SALS under H_V (cross‐polarized) optical alignments during melting can be used to reliably measure the thermodynamic (e.g., crystal melting and melt‐phase separation temperatures) and structural variables (e.g., crystalline fraction within the superstructures and volume fraction of superstructures) necessary for describing the multiphase behavior of crystal‐amorphous blends in one combined measurement. We also evaluate the orientation correlations of crystalline volume elements within the superstructures. Our results indicate that simultaneous measurement of transmitted light can provide a reliable estimate of the total scattering from density and orientation fluctuations and the melt‐phase separation temperature of polymer blends. For solution‐cast poly(?‐caprolactone)/poly(D,L‐lactic acid) blends, our multivariable measurements during melting provide the parameters necessary to generate a crystal–liquid and liquid–liquid phase diagram and characterize the solid‐state morphology. This opens up the challenge to explore use of our vertical SALS instrument as a rapid and convenient method for developing structure–property relationships for crystal‐amorphous polymer blends. © 2002 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 40: 2714–2727, 2002     

Miscibility and enzymatic degradation studies of poly(ε-caprolactone)/poly(propylene succinate) blends

In the present study the miscibility behaviour and the biodegradability of poly(ε-caprolactone)/poly(propylene succinate) (PCL/PPSu) blends were investigated. Both of these aliphatic polyesters were laboratory synthesized. For the polymer characterization DSC, ¹H NMR, WAXD and molecular weight measurements were performed. Blends of the polymers with compositions 90/10, 80/20, 70/30 and 60/40 w/w were prepared by solution-casting. DSC analysis of the prepared blends indicated only a very limited miscibility in the melt phase since the polymer-polymer interaction parameter χ₁₂ was −0.11. In the case of crystallized specimens two distinct phases existed in all studied compositions as it was found by SEM micrographs and the particle size distribution of PPSu dispersed phase increased with increasing PPSu content. Enzymatic hydrolysis for several days of the prepared blends was performed using Rhizopus delemar lipase at pH 7.2 and 30 °C. SEM micrographs of thin film surfaces revealed that hydrolysis affected mainly the PPSu polymer as well as the amorphous phase of PCL. For all polymer blends an increase of the melting temperatures and the heat of fusions was recorded after the hydrolysis. The biodegradation rates as expressed in terms of weight loss were faster for the blends with higher PPSu content. Finally, a simple theoretical kinetic model was developed to describe the enzymatic hydrolysis of the blends and the Michaelis-Menten parameters were estimated.     

The assessment of miscibility and morphology of poly(ε-caprolactone) and poly(para-chlorostyrene) blends

The miscibility and morphology of poly(ε-caprolactone) (PCl) and poly(para-chlorostyrene) (PpClS) blend were investigated by using thermal analysis, morphological analysis, viscometry, and the study of melting point depression. A single glass transition temperature was observed by differential scanning calorimetry (DSC) for PCl/PpClS blends in the whole compositional range (0/100, 25/75, 50/50, 62.5/37.5, 75/25, 90/10). Morphology of the polymers and their blends was studied by scanning electron microscopy (SEM). The Fourier transform infrared spectra of the samples were obtained by spectrometer. Up to 12 cm⁻¹ shifts in carbonyl stretching band of PCl was detected in the spectra of PpClS rich blends. The viscosity of PCl, PpClS and their blends has also been studied to investigate the miscibility according the miscibility criteria Δb, and Δ[η]. Using this data, the interaction parameters α and μ, based on the Chee and Sun et al. approaches were determined. These criteria indicated that the blend is miscible in all proportions up to 90% of PCl content in the blends. The melting point depression of PCl in the blends was examined to obtain the interaction parameter, χ₁₂ for this system. The parameter, χ₁₂ was found to be composition dependent. Negative values of the obtained interaction parameter also support the miscibility of this system up to the 90% PCl in the blend.     

Compatibility study of quetiapine fumarate with widely used sustained release excipients

The drug-excipient compatibility study of quetiapine fumarate, with widely used sustained release excipients, was carried out employing differential scanning calorimetry (DSC) and Fourier transform infrared spectroscopy (FT-IR). The selected excipients were HPMC K100M, sodium alginate, xanthan gum, Eudragit RSPO, hydrogenated castor oil, carnauba wax, and PEO WSR 303. Equal proportion of drug and excipients was utilized in the interaction study. FT-IR spectra indicated the absence of interaction between drug and excipients. The DSC curve showed a sharp endothermic melting peak at 173.26 °C for quetiapine fumarate. Post melting interaction was observed for carnauba wax, Eudragit RSPO, and hydrogenated castor oil probably due to solubilization of drug in the melted excipient. No interaction was observed for other excipients. The physical mixtures stored at 30 ± 2 °C/65 ± 5% RH did not show any significant degradation of the drug. The concept of systemically conducted preformulation studies will facilitate dossier submission to the drug control authority.     

Cure behavior and thermal stability of an epoxy: new bismaleimide blends for composite matrices

A new bismaleimide (BMI) resin was synthesized to formulate epoxy(tetraglycidyl diaminodiphenyl methane; TGDDM) – bismaleimide thermoset blends for composite matrix applications. 4,4′-diaminodiphenyl methane (DDM) was used as an amine curing agent for the TGDDM. A Fourier transform infrared (FTIR) spectroscopy was employed to characterize the new BMI resin. Cure behavior of the epoxy–BMI blends was studied using a differential scanning calorimeter (DSC). DSC thermograms of the thermoset blends indicated two exothermic peaks. The glass transition temperature of the thermoset blends decreased with BMI content. Thermogravimetric analysis (TGA) was carried out to investigate thermal degradation behavior of the cured epoxy–BMI thermoset blends. The new BMI resin reacted partially with the DDM and weak intercrosslinking polymer networks were formed during cure of the thermoset blends.     

Biodegradable poly(alkylene succinate) blends: Thermal behavior and miscibility study

New binary blends composed of poly(ethylene succinate) and poly(propylene succinate) or poly(ethylene succinate) and poly(butylene succinate) were prepared. Both PESu/PPSu and PESu/PBSu systems belong to semicrystalline/semicrystalline pairs. The miscibility and crystallization behavior was investigated using differential scanning calorimetry (DSC), wide angle X‐ray diffraction (WAXD), and polarizing light microscopy (PLM). Blends of PESu and PPSu exhibited a single composition dependent glass transition temperature over the entire range of composition, indicating that the system is miscible. The melting point depression of the high melting temperature component, PESu, was analyzed according to the Nishi‐Wang equation. A negative polymer–polymer interaction parameter was obtained, indicating that the blends are thermodynamically miscible in the melt. The two components crystallized sequentially when the blends were cooled rapidly to a low temperature. DSC traces of PESu/PBSu blends after quenching showed two distinct composition dependent glass transition temperatures between those of the neat polymers, showing that the polymers are partially miscible. The amorphous PESu/PBSu blends in the intermediate compositions showed three cold‐crystallization peaks, indicating the influence of mixing. The crystallization rates of PBSu were reduced and those of PESu were increased. WAXD showed reduced crystallinity and peak broadening in the patterns of the blends of intermediate compositions, while no spherulites could be detected by PLM. © 2005 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 44: 584–597, 2006     

Miscibility and thermal stability of ethyl vinyl acetate and ethylene-octane copolymer blends

Miscibility and thermal stability of ethyl vinyl acetate (EVA) and ethylene octane (EO) copolymer blends with different compositions were investigated by thermogravimetric analysis (TGA), differential scanning calorimetry (DSC) and atomic force microscopy (AFM). The degradation behavior by TGA data under dynamic conditions in an inert atmosphere shows the blends to be immiscible. The addition of EO improves the thermal stability of EVA for all composition and temperature ranges. Using the DSC experiment, two single crystallization temperatures (T _c) for the blends were obtained and the crystallization and melting enthalpy with compositions abiding by the additive rules, confirm the immiscibility of the blends. The rate of crystallization seems to be independent of blend compositions. The surface morphology using AFM shows a thin and elongated crystallites of pure EO, and a bulky and random morphology for EVA, where a perfect mixture of aforementioned structures for 50/50 blend, with the immiscible domains of both EO and EVA. The 2D-power spectral density (PSD) analysis shows the surface roughness of 50/50 blends is in between of EO and EVA. Both AFM observations and quantitative PSD results, along the line with DSC and TGA. The experimental data for miscibility and stability by TGA, DSC and AFM techniques reveal that blends of EVA/EO are immiscible in the entire range of compositions.     

The miscibility of poly(vinyl alcohol)/poly(N-vinylpyrrolidone) blends investigated in dilute solutions and solids

Poly(vinyl alcohol) (PVA) (polymer A) and poly(N-vinylpyrrolidone) (PVP) (polymer B) are known to form a thermodynamically miscible pair. In the present study the conclusion on miscibility of PVA/PVP solid blends, confirmed qualitatively (DMTA, FTIR) and quantitatively (DSC, χ_AB = − 0.69 at 503 K) is compared with the miscibility investigations of PVA/PVP solution blends by the technique of dilute solution viscometry. The miscibility of the ternary (polymer A/ polymer B/ solvent) system is estimated on the basis of experimental and ideal values of the viscosity parameters k, b and [η]. It is found that the conclusions on miscibility or nonmiscibility drawn from viscosity measurements in dilute solution blends depend: (i) on the applied extrapolation method used for the determination of the viscosity interaction parameters, (ii) on the assumed definition of the ideal values and (iii) on the thermodynamic quality of the solvent, which in the case of PVA depends on its degree of hydrolysis. Hence, viscometric investigations of dilute PVA/PVP solution blends have revealed that viscometry, widely used in the literature for estimation of polymer-polymer miscibility can not be recommended as a sole method to presume the miscibility of a polymer pair.     

聚乳酸/羧基化聚丙烯共混物的形态与热性能研究

以扫描电子显微镜、热重分析仪、差示扫描量热仪、热台偏光显微镜分别研究了聚乳酸/羧基化聚丙烯共混体系的相形态、热性能和结晶形态.结果显示,共混物熔体冷却时,聚乳酸和羧基化聚丙烯均形成球晶,但羧基化聚丙烯球晶较大而十字消光较暗,聚乳酸球晶尺寸较小而十字消光较亮,且聚乳酸球晶产生规则的、不连续的同心环线——裂纹,裂纹厚度约为1~2μm,且裂纹内部有微纤存在.当聚乳酸含量≤50%时,由于聚丙烯上羧基的存在而使共混体系具有较好的相容性.共混物的热分解过程分为三个阶段,热分解温度的变化是聚丙烯上的羧基、聚乳酸和聚丙烯骨架分解三种机制共同作用的结果,少量聚乳酸能够明显提高共混物中聚丙烯上羧基的热稳定性.共混物中的羧基化聚丙烯组分可以发挥稀释剂的作用,大幅度降低了聚乳酸的冷结晶温度.聚乳酸含量≥50%时,共混熔体降温时DSC谱图中聚乳酸和羧基化聚丙烯分别结晶,而聚乳酸含量<50%时,只观察到羧基化聚丙烯的结晶行为.     

Die ermittlung des schmelzpunktes von kristallinen polymeren mittels wärmeflusskalorimetrie (DSC)

Small and imperfect crystals in polymers reorganize during slow heating. This leads to an increase of their melting point T_m. In order to measure the melting point of the original crystals, high heating rates are needed. This is possible with the modern heat-flow-calorimeters, which work with very small samples. The thermal lag of a DSC cell causes a shift of the melting peak by ΔT to higher temperature. From the theory of a heat-flow-calorimeter, it follows that the error ΔT is proportional to the square root of heating rate. heat of fusion and sample mass. Measurements with sharp melting low molecular weight compounds confirm that this square root relation is quantitatively followed. In order to measure the true melting point of the crystals present in a polymer sample, one has to use different high heating rates and constant sample mass. By plotting the melting peak temperatures as a function of the square root of heating rate and linear extrapolation to zero heating rate, the true melting point is found. This method is applied to HDPE, LDPE and some polyamides.     

Studies on poly(ε‐caprolactone)/thiodiphenol blends: The specific interaction and the thermal and dynamic mechanical properties

The effects of several low molecular weight compounds with hydroxyl groups on the physical properties of poly(ε‐caprolactone) (PCL) were investigated by Fourier transform infrared (FTIR) spectroscopy and high‐resolution solid‐state ¹³C NMR. PCL and 4,4′‐thiodiphenol (TDP) interact through strong intermolecular hydrogen bonds and form hydrogen‐bonded networks in the blends at an appropriate TDP content. The thermal and dynamic mechanical properties of PCL/TDP blends were investigated by differential scanning calorimetry (DSC) and dynamic mechanical thermal analysis, respectively. The melting point of PCL decreased, whereas both the glass‐transition temperature and the loss tangent tan δ of the blend increased with an increase in TDP content. The addition of 40 wt % TDP changed PCL from a semicrystalline polymer in the pure state to a fully amorphous elastomer. The molecules of TDP lost their crystallizability in the blends with TDP contents not greater than 40 wt %. In addition to TDP, three other PCL blend systems with low molecular weight additives containing two hydroxyl groups, 1,4‐dihydroxybenzene, 1,4‐di‐(2‐hydroxyethoxy) benzene, and 1,6‐hexanediol, were also investigated with FTIR and DSC, and the effects of the chemical structure of the additives on the morphology and thermal properties are discussed. © 2000 John Wiley & Sons, Inc. J Polym Sci B: Polym Phys 38: 1848–1859, 2000     

Thermal and dynamic mechanical properties of PES/PPS blends

Blends of poly(ether-sulfone) (PES) and poly(phenylene sulfide) (PPS) with various compositions were prepared using an internal mixer at 290°C and 50 rpm for 10 min. The thermal and dynamic mechanical properties of PES/PPS blends have been investigated by means of DSC and DMA. The blends showed two glass transition temperatures corresponding to PPS-rich and PES-rich phases. Both of them decreased obviously for the blends with PES matrix. On the other hand, T_g of PPS and PES phase decreased a little when PPS is the continuous phase. In the blends quenched from molten state the cold crystallization temperature of PPS was detected in the blends of PES/PPS with mass ratio 50/50 and 60/40. The melting point, crystallization temperature and the crystallinity of blended PPS were nearly unaffected when the mass ratio of PES was less than 60%, however, when the amount of PES is over 60% in the blends, the crystallization of PPS chains was hindered. The thermal and the dynamic mechanical properties of the PPS/PES blends were mainly controlled by the continued phase. This revised version was published online in August 2006 with corrections to the Cover Date.     

Flory-huggins interaction parameters of LCP/thermoplastic blends measured by DSC analysis

Liquid crystalline polymer/polyamide 66 (LCP/PA66) and LCP/poly(butyl terephthalate) (LCP/PBT) blends were compounded using a Brabender Plasticorder equipped with a mixing chamber. The LCP employed was a semi-flexible liquid crystalline copolyesteramide based on 30 mol% of p-amino benzoic acid (ABA) and 70 mol% of poly(ethylene terephthalate) (PET). The Flory-Huggins interaction parameters (χ12) of the LCP/ PA66 and LCP/PBT blends are estimated by melting point depression from DSC measurement. The results indicate that c12 values all are negative for LCP/PA66 and LCP/PBT blends, and when the LCP content in these blends is more than 10 mass%, the absolute value of χ₁₂ decreases. Thereby, we can conclude that LCP/PA66 and LCP/PBT blends are fully miscible in the molten state, the molecular interaction between the LCP and PA66 is stronger than that between LCP and PBT. As the LCP content in LCP/PA66 and LCP/PBT blends is more than 10 mass%, the molecular interaction between LCP and matrix polymer decreases. This revised version was published online in August 2006 with corrections to the Cover Date.     

Study of the miscibility and the thermal degradation of PVC/PMMA blends

The aim of this paper is to study the miscibility and the thermal degradation of PVC/PMMA blends. For that purpose, blends of variable compositions from 0 to 100 wt% were prepared with and without plasticizer. Their physico-chemical characterization was carried out by differential scanning calorimetric analysis (DSC) and Fourier transform infrared spectroscopy (FTIR). Their thermal degradation under nitrogen at 185°C was studied and the HCl evolved from PVC was measured by the pH method. Degraded samples were characterized, after purification, by FTIR and UV-visible spectroscopy. The DSC analysis showed polymer miscibility up to 60 wt% of PMMA. This miscibility is due to a specific interaction of hydrogen bonding type between carbonyl groups (C=O) of PMMA and hydrogen from (CHCl) groups of PVC as evidenced by FTIR analysis. On the other hand, it was found that PMMA exerted a stabilizing effect on the thermal degradation of PVC by reducing the zip dehydrochlorination and by leading to the formation of short polyenes.