When dealing with challenges such as providing fire/flame protection in combination with weathering resistance to polymers, one needs to take into account that the flame retardant might degrade independently (e.g., aromatic brominated flame retardants are not UV stable and release bromine and aromatic radicals that initiate photo-oxidation of the host polymer). Also, traces of metal impurities in filler-based flame retardant systems function as prodegradants by promoting hydroperoxide decomposition to free radical products that cause premature polymer failure. In addition, ammonium-modified clay nanocomposites are prone to Hofmann elimination reactions during polymer processing whereby flammable and volatile olefins and amines will be released. Phosphorus-based flame retardants may hydrolyze during polymer weathering whereby polyphosphoric acids decrease HALS activity through protonation. There might also be a direct interaction between the polymer and the flame retardant owing to poor compatibility and solubility or an interaction between the flame retardant and the stabilizer additives. For instance, the inclusion of filler-based flame retardants may lead to adsorption/desorption of stabilizers and antioxidants on the filler surface whereby the thermal and photostability of the polymer may be drastically reduced (the host polymer will be depleted of stabilizer owing to stabilizer sorption onto the high energy surface of various flame retardant fillers) or in the case of aromatic brominated flame retardants the formed acidic vapors may deactivate the HALS through ammonium salt formation whereby again the polymer service life will be significantly reduced as shown in Figure 3.
Thus, certain flame retardant types may cause difficulties in regard to the durability of polymeric materials as a result of their reactivity and crossreactivity with other additives such as stabilizers and fillers. Overall, the demands on flame-retarded products and their performance are becoming ever more stringent owing to higher service-life expectations, new more challenging applications and standards, tougher legislation and regulations, lower cost requirements, and above all the request for flame retardant formulations with better toxicological and environmental profiles.
Brominated Flame Retardants and Their Impact on Polymer Stability
The mode of action of brominated flame retardants is based on thermally induced release of bromine radicals at the site of the flame zone. It has been established that the relative efficacy of halogenated flame retardants increases with decreasing thermal stability. However, a drawback of conventional brominated flame retardants such as tetrabromobisphenol A or brominated diphenyl ethers, besides environmental concerns, is that they have a tendency to undergo dehydrobromination reactions already during polymer processing (temperature range, 210–250°C) and when subjected to UV irradiation during end use. The dehydrobromination reaction involves the homolytic cleavage of the CBr bond, leading to the formation of bromine and aromatic radicals that will induce photolytic and thermo-oxidative reactions in the polymer matrix, that is thereby accelerating the degradation by “feeding” the auto-oxidation cycle with radicals. For polypropylene, it has been observed that the rate of formation of photoproducts is dramatically enhanced in the presence of decabromodiphenylether and antimony oxide compared to a virgin polypropylene sample.20–22 Similar results have also been recorded for halogen flame-retarded polystyrene21 and polyethylene (HDPE, LDPE).23
In addition, hydrobromic acid is formed when the bromine radicals abstract hydrogen from the polymer backbone itself. The formed halogenated acid vapors are in turn known to impair the performance of HALS additives through the formation of an amine salt through protonation of the amine.24 The amine salt inhibits the oxidation of HALS to the nitroxyl radical that is mandatory for polymer stabilization to take place. On top of this, the resulting amine salt is less thermally stable than the corresponding amine and it may even lead to the total decomposition of the HALS skeleton during processing.25 Aliphatic bromine compounds are, in this respect, less detrimental than aromatic ones.26 Studies have shown that even very weak acids such as those generated from sterically hindered phenolic stabilizers27 may partially deactivate the HALS and that also bromine radicals may react with the nitroxyl radical of the 2,2,6,6-tetramethylpiperidine light stabilizer.28
Several strategies and attempts have been made to circumvent the antagonistic effect between brominated flame retardants and HALS derivatives to secure high weathering resistance of halogen flame-retarded polymers. Mainly, four different strategies have been tested: (1) the use of acid scavengers such as tin maleates or antiacids such as ZnO, calcium stearate, and Mg(OH)229; (2) the use of traditional UV absorbers or pigments as UV filters30; (3) microencapsulation of the brominated flame retardant31; and (4) utilization of low interacting HALS derivatives where the HALS moiety has been replaced with N-alkoxy or N-acyloxy hindered amines.32 The most successful approach of these aforementioned strategies has been the development of alkoxyamine (NOR) derivatives that exhibit lower basicity compared to conventional HALS and which are already in a more active oxidation state than classical HALS derivatives for polymer stabilization.33–35 In addition, good results have been obtained by combining NOR compounds with UVA. The primary benefit of UVA is that it provides an UV screen for the aromatic brominated flame retardant besides the polymer itself, thus inhibiting the generation of bromine radicals and hydrobromic acid.30 Further study has shown that UL94 V-0 rating can be achieved with NOR and a halogenated system.19 Thus, the NOR additives do not only improve the UV stability and long-term stability of polymers, but also exhibit a strong synergistic effect with brominated flame retardants. As an example, a commercial multicomponent system consisting of brominated flame retardant/antimony oxide, NOR light stabilizer, and selected pigment has been successfully developed for stadium seats made out of polypropylene. This FR/NOR light stabilizer formulation for PP copolymers survives 4000 h of artificial weathering (ASTM G26, spray), whereas the similar formulation with conventional FR/HALS light stabilizer combination fails already after about 800 h.35 In addition, it has been proposed that the detrimental interaction between brominated flame retardant and HALS can be partially circumvented by microencapsulation of the halogenated flame retardant.36
Phosphorus- and Nitrogen-Based Halogen-Free Flame Retardants
Ammonium polyphosphate (APP), which is available in different crystal modifications and with coatings (e.g., melamine and silicone) or in an encapsulated form, decomposes by exposition to fire or heat to ammonia, polyphosphoric acid, and phosphorus oxides. In the presence of a char-forming agent, for example a polyol, an intumescent barrier via phosphoric ester intermediates is formed. APP-/pentaerythritol-based intumescent flame retardants do not significantly accelerate the degradation of unstabilized PP.37, 38 However, the APP-based intumescent flame retardant is susceptible to photodegradation as such. Therefore, it can be concluded that two separate processes take place, that is (1) photo-oxidation of the intumescent flame retardant and (2) photodegradation of the polymer matrix. Other studies have shown that the activity of HALS is somewhat reduced likely through protonation by the polyphosphoric acid. It is, however, less severe than with brominated flame retardants. On the other hand, the efficacy of an UV absorber is increased through the polyol present in the flame retardant combination.
The photochemical behavior of polypropylene formulations containing 0.5 wt % HALS as light stabilizer and as flame retardant either octabromodiphenyl ether (OBDE)/Sb2O3 or APP has been studied under artificial accelerated conditions.39 The formulation with OBDE showed significant surface degradation (the appearance of surface cracks by light microscopy at 100-fold magnification was used as the criteria of surface degradation) already after 400 h irrespective of the presence of 0.5 wt % of HALS, whereas the APP/HALS system was stable up to 1800 h of artificial light exposure. On the other hand, the reference polypropylene/HALS sample without any flame retardant lasted for 2500 h under similar experimental conditions. However, in case of artificial weathering, hydrolysis of APP may occur and it results in reduced fire retardancy as the formation of the intumescent network is disturbed already after 210 MJ/m2, corresponding to 1 year in mid-Europe.40 A similar result was found in EVA/PA-6/PP-blends.41
As red phosphorus is mainly used in engineering plastics (polycarbonate, polyamide, and polyester), the influence on light- and long-term thermal stability is less critical. From the stabilization point of view, it is important to stabilize red phosphorus against phosphine formation through moisture which is achieved by microencapsulation42 or by adding salts such as copper acetate.43
Nitrogen-containing flame retardants are a class of various materials among which are well-known halogen-free commercial products, namely APP, melamine cyanurate (MC), melamine borate, and melamine (poly)phosphate. MC is a 1 : 1 adduct of melamine and cyanuric acid. MC decomposes endothermically into its components and melamine decomposes further to nitrogen-containing gases such as noncombustible ammonia. Stabilized polyamide MC formulations did not lose their fire retardancy after artificial weathering (210 kJ/m2); however, some blooming of MC was detected.40 Melamine phosphates did not reveal any antagonistic effect with HALS in artificial weathering experiments.44 Actually, even a slight improvement could be found owing to a pigment effect of the flame retardant.
The processing stability of MC-containing polyamide has been shown to be improved by addition of sulfates (e.g., magnesiumsulfate45) or metal acetates (e.g., magnesium acetate46). Bisphenol-A-diphosphate PC/ABS blend showed without UV stabilizer discoloration and surface cracks, it had, however, no influence on the flame retardancy after 210 MJ/m2.40
MC polyamide compounds, APP polypropylene, and melamine phosphate/polypropylene can be successfully recycled without significant loss of properties.47 On the contrary, recycled PC/ABS flame retarded with organic phosphate ester lost already after the first recycling step flame retardancy, whereas its molecular weight and mechanical properties were still essentially unchanged.48 In contrast, ABS with brominated flame retardants passed the UL 94 5VB test even after five extrusion steps.
Nitrogen Flame Retardants Based on Dual Functional NOR Additives
It was earlier believed that the noninteractive NOR additive in the presence of brominated flame retardants contributed only to improved light stability and that it functioned as an FR synergist with halogenated flame retardants. More recently, it has been shown that in some polypropylene formulation, the NOR additive can successfully replace antimony oxide as a synergist for halogenated flame retardants. Later, it was discovered that NOR additives, in fact, exhibit flame retardant and self-extinguishing properties in polypropylene films and fibres by themselves.49–52 Other advantages of NOR additives are that they are easily melt processable, they function at very low concentrations (ca. 1 wt %), and they do not weaken the mechanical or physical properties of the host polymer. In short, NOR provide flame retardancy in thin polypropylene/polyolefine sections in combination with inherent light stability properties. A disadvantage is that the NOR compounds alone do not achieve usual flame retardant standards such as UL 94 in polypropylene moldings and that they cannot be processed above 250°C without decomposition. In addition, several other chemical structures related to alkoxyamines are claimed to provide flame retardancy namely hydroxylamine esters,53 hydroxylamines, and nitroxyl radicals.
The activity of alkoxyamines as flame retardants is based on the thermolysis of nitroxyl ethers which leads to the formation of either alkoxy and aminyl radicals or alkyl and nitroxyl radicals (Figure 4).
However, the literature available on the role of various NOR structures on fire retardant efficacy is still very limited and mostly in the form of patents. A few examples of structure–property relationship of various NOR structures in flame-retarded polypropylene films are shown in Figure 5.
Figure 5. Flame retardant ranking of various NOR additives based on their performance in DIN 4102/B2 test in polypropylene films.
Download figure to PowerPoint
In the series of 1-n-alkoxy-4-(dodec-2enyl)-2,2,6,6-tetramethyl-piperidines, all of the compounds showed comparable or even higher efficiency than FLAMESTAB NOR116 (commercial product, supplied by BASF SE, structure attachment). The best flame retardant effect in this series was obtained for 1-methoxy-4-(dodec-2enyl)-2,2,6,6,tetramethyl-piperdine. The results indicate that the flame retardant properties within the family of NOR compounds increased as a function of the thermal stability of the N-alkoxyamine additive.55 Similar ranking results for NOR efficacy as flame retardants can also be found in a recent patent application for new spiro-NOR-HALS derivatives.56
Novel Multifunctional (1-Alkoxy-2,2,6,6-tetramethylpiperidin-4-yl)diazene Additives
Recently, it has been reported that (1-alkoxy-2,2,6,6-tetramethylpiperidin-4-yl)diazene (AZONOR) compounds provide, besides flame retardancy also, heat and light stability to polypropylene films and plaques.57–59 The AZONOR compounds are active at very low loadings of only 0.5 wt % and they have no detrimental effect on polypropylene appearance or its mechanical and processing properties. Moreover, they exhibit a synergistic effect with many conventional flame retardants such as brominated, inorganic, and especially phosphorous flame retardants.60 Representative AZONOR structures are shown in Figure 6.
In the shown experiment, the various AZONOR compounds were the only ones at the concentration applied to provide flame retardancy in polypropylene plaques of 1 mm for more than 2000 h of artificial weathering (WOM Ci 65A, BPT 63°C, 60% relative humidity, water spray) as summarized in Table I. Another bonus of AZONOR flame retardants was that no burning dripping could be detected.
Table I. Results of Flame Retardant Testing According to DIN4102/B2 Standard Before and After Artificial Weathering
| ||Before artificial weathering||After 2000 h of artificial weathering|
|Additives (wt %)||Weight loss (%)||Burn length (mm)||DIN 4102 Pass/fail||Weight loss (%)||Burn length (mm)||DIN 4102 Pass/fail|
|Blank||100||100||Fail||Test failed before WOM test|
|NOR371(2%)||n.d.||88||Fail||Test failed before WOM testa)|
|NOR116 (0.5%)||49.4||80||Fail||Test failed before WOM test|
|1 (0.5%)||n.d||31||Pass||Passed 2000 h of WOM|
|2 (0.5%)||n.d||41||Pass||Passed 2000 h of WOM|
|3 (0.5%)||5.2||27||Pass||Passed 2000 h of WOM|
|4 (0.5%)||14||43||Pass||Passed 2000 h of WOM|
|5 (0.5%)||8.9||37||Pass||Passed 2000 h of WOM|
Fillers and Polymer Stability
The impact of various inorganic fillers on the stability of polymers is illustrated in the following examples: The long-term heat stability at 150°C in a forced air circulated oven for three talc-filled polypropylene grades decreased from 80 days (unfilled polypropylene, base stabilization package: 0.1 calcium stearate, 0.1 wt % AO-1, and 0.3 wt % PS-1 to 5–24 days) until embrittlement depending on the specific talc used. Thus, the contribution of talc to heat stability was always negative, even if the amplitude was dependent on talc type.61 Similar results of decreased stability of talc and calcium carbonate-filled polypropylene composites have also been observed by other research groups.62–64 Experiments using talc-filled polypropylene samples containing a 1 : 1 mix of LS-1 and LS-2 as light stabilizers have also been carried out. The results from the artificial weathering experiments clearly show that talc also significantly decreases the light stability of polypropylene. The observed decrease in both heat and light stability has been mainly attributed to stabilizer absorption on the high energy surface of the filler. In addition, it was showed that heat and light stability of the composites could be improved by using a “filler deactivator” such as an oligomeric Bisphenol-A-glycidylether or by significantly increasing the amount of light stabilizer. The increase of stabilizer content may not be a feasible, economic, or practical solution to the stabilizer adsorption problem, whereas the filler deactivator route holds more promise. For instance, a number of filler deactivators or coupling agents have been successfully used to modify the filler surface and to improve the stability of the polymer. Suitable additives range from typical filler coatings (stearic acid, stearates), oligomeric epoxides, silanes, titanates, to functional polymers (e.g., polypropylene-graft-maleic anhydride or polypropylene-graft-acrylic acid). Encouraging results for using coupling agents with respect to improved heat stability of talc-filled polypropylene formulations have been recorded, that is certain amphiphilic coupling agents with hydrophilic ends prevent the undesired adsorption of stabilizers onto the filler surface.65, 66
Nanocomposites—Highly Dispersed Flame Retardants
In recent years, fillers in nanodimensions such as highly dispersed flame retardants, layered and fibrillated silicate clays, carbon nanotubes and nanofibers, calcium carbonate, metal oxides, or silica nanoparticles that function at low loadings of 2–5 wt % have been investigated. Researchers have focused their attention particularly on polymer-layered silicate nanocomposites. The clay-based nanocomposites are among the most examined owing to the relatively low price of clay minerals, their availability, and unique characteristics including their plate-like morphology with a high aspect ratio. The macroscopic phase behavior of polymer–colloid mixtures has been the subject of many theoretical and experimental studies. The dispersion state of the nanoclay in the polymer matrix can be described by cluster, intercalated (distanced but yet parallel layers), or exfoliated (fully dispersed) structures. In each case, the polymer chains are differently arranged in the nanocomposite matrix and thus the composite properties are intimately linked to the macroscopic structure of the inorganic filler in the nanocomposite. In many cases, the most significant improvements in reinforcement, barrier, and flame retardant properties have been found for intercalated and/or exfoliated systems and therefore various ways to enhance the dispersibility have been the topic of great interest67–72 Normally, polymers containing a small amount of nanofiller exhibit a significant decrease in peak heat release and mass loss during cone calorimeter experiments. The fire retardant effect has been attributed to the accumulation of inorganic particles at the surface of the composite with subsequent formation of a carbonaceous coating limiting heat and mass transfer. However, nanofillers rarely contribute, in their own right, to improvements in traditional fire tests such as UL-94 or LOI tests. Today, there is a consensus that nanofillers need a synergistic flame retardant to perform well in the aforementioned two fire tests. Recently, it has also been noted that nanocomposites prevent blooming by reducing environmental release of any additive present in the nanocomposite product.
The influence of nanosized silica on the oxidative and photo-oxidative stability has been studied by several authors. The results vary somewhat depending on the silica used in terms of particle size, pore, and surface morphology, purity in terms of metal residues (Ti, Al, Fe), Bronsted and Lewis acid sites, and controlled antioxidant absorption/desorption characteristics. It has been shown that various antioxidants are absorbed on the silica surface with different strengths depending on their chemical structure, that is studies have shown that light stabilizers, for example Chimasorb 944 (HALS-x) are more strongly adsorbed on the silica surface than hindered phenols such as butylated hydroxyl toluene (BHT) or (AO-1). When either of the antioxidants was used as a single stabilizer together with silica in LDPE, a reduction of stabilization performance was recorded. This phenomena was attributed to the deactivation of the antioxidant through immobilization of the active groups of stabilizers molecules such as OH, >NH, or >NO at the filler surface. Although interestingly it was noticed that when both HALS and hindered phenol additives were present in the silica/LDPE formulation, a strong synergistic effect was observed, that is leading to an enhanced level of stabilization. This in turn was ascribed to controlled release of the phenolic antioxidant through displacement from the silica surface by the more strongly adsorbing HALS.73 Another study revealed that metal ion (Al, Ti, Fe) impurities in silica significantly accelerate the degradation process by catalyzing the breakdown of formed hydroperoxides at already very low loadings of silica. Studies have shown that the drawback of metal ion contaminants can be circumvented by using appropriate metal deactivators such as 1,2-bis(3,5-di-tert-butyl-4-hydroxyhydro-cinamoyl)hydrazine (MD-1).
As mentioned earlier, layered silicates have been evaluated as cost effective, environmentally friendly additives for reducing flammability, and for improving physical properties of various polymers.74–83 The clay-based nanocomposites are among the most examined owing to the relatively low price of clay minerals, their availability, and unique characteristics including their plate-like morphology with a high aspect ratio. Again, layered silicates from natural sources (e.g., montmorillonite) may contain evenly distributed metal ions as contaminants that are known to have a negative effect on polymer stability. Polymer compatibility with the surface of the clay platelets is crucial for obtaining sufficient dispersion of the nanoparticles in the polymer matrix. For instance, Seo et al.84 prepared polyurethane nanocomposites by utilizing silanol surface-modified clays that reacted with NCO groups of polymeric 4,4′-diphenyl methane diisocyanate (MDI) whereby an exfoliated structure was obtained that resulted in enhanced mechanical properties at a clay loading of 3 wt %. Camino and coworkers85 have shown that in the case of modified nanodispersed clay in polyurethane formulations, improved flame retardant properties can be reached for both intercalated and exfoliated structures. In general, it is difficult to obtain a homogeneous dispersion of nanoparticles in a polymer by using existing/traditional compounding techniques owing to the strong tendency of fine (nano) particles to agglomerate. One widely explored route to gain better interaction between the clay–polymer interfaces is to modify the nanoparticle surface by treatment with surfactants such as alkyl ammonium surfactants with long alkyl tails. In this case, the amount of organic material within the clay can be very high (ca. 40%), and therefore, the stability of this organic part of the surfactant cannot be neglected. At normal polymer processing temperatures of above 200°C, the thermal stability of the ammonium salts is too low. Most of the ammonium structures tend to undergo Hofmann elimination reactions whereby volatile and flammable olefin and amine derivatives are released. In fact, the thermal degradation of ammonium salts starts already at 180°C and is furthermore reduced by catalytically active sites on the alumosilicate layer.69 Hydrophobic polymers such as polyolefins often require, on top of the organic modification filler, a substantial amount of an additional compatibilizer such as polypropylene-g-maleic anhydride (PP-g-MAH). These lower molecular weight compatibilizers usually exhibit an inferior oxidative stability compared to the parent polymer and, therefore, reduce the long-term performance of polymer nanocomposites.74
In many cases, nanocomposites based on organically modified montmorillonite show in comparison to neat polymer a dramatic reduction of long-term stability,71–73 for example the long-term thermal stability of a stabilized polypropylene at 135°C is reduced from more than 40 days to only 15 days in the presence of 5% organically modified clay. Noteworthy is that the nanocomposites' higher propensity to photo-oxidation also leads to faster discoloration and deterioration of mechanical properties. For instance, according to the artificial weathering experiments, polypropylene montmorillonite-based nanocomposites (stabilized with antioxidant) degrade under exposure to UV light much more rapidly than virgin polypropylene.81, 86 The photo-oxidation of EPDM montmorillonite nanocomposites drastically reduced the induction time of photo-oxidation in the presence of the nanoclay and the effect was enhanced in the presence of an exfoliated nanocomposite structure.82 The photo-oxidation of syndiotactic PP/synthetic clay (fluorohectorite modified by octadecylammonium) nanocomposites showed that the nanoparticles were catalyzed the decomposition process.83 The presence of PP-g-MAH as compatibilizer accelerated the degradation furthermore. In natural clays, iron impurities play an active role in the dramatic modification of the oxidation kinetics.74 Similar results of fast photo-oxidative degradation have also been found for polyethylene/montmorillonite nanocomposites. The reason for the more rapid degradation has been attributed to the adverse role played by the thermal decomposition products of the alkyl ammonium surfactant (Hofmann elimination reaction), the photo instability of compatibilizers such as PE-g-MA, and negative contribution from exfoliated structures.87–89 The influence of the nanocomposite substrate is furthermore of importance, for example EVA or polystyrene shows less acceleration of degradation than LDPE.90–94 In addition, for nanodispersed hydrotalcite in polypropylene95 for CaCO3 and SiO2 nanoparticles,96 for EPDM nanocomposites synthesized from layered double hydroxides,97 and for boehmite, modified by long-chain alkyl benzene sulfonic acid in isotactic PP and in syndiotactic PP98 a similar phenomenon of accelerated photo-oxidation has been recorded.
Only in some rare cases such as polyamide-693 and PE/PA-6 blends,94 an improvement in (photo)oxidative stability in the presence of nanoclays could be detected.
Overall, it seems that in the case of filled polymers that the dilemma of the adsorption of stabilizers (in particularly for exfoliated structures) and the level of metal impurities need to be carefully addressed to achieve sufficient long-term heat and light stability of flame-retarded products. The conventional approach of melt blending of stabilizers into the nanocomposite during their production is insufficient. Thus, for example PP stabilized with 0.05% phenolic antioxidant and 0.05% phosphite achieves only a life time of 1.8 min (fluorohectorite) or 2.3 min (montmorillonite) at 190°C. However, through proper stabilization including filler deactivators, the oxygen induction time (OIT) value can be prolonged to more than 90 min.99 With similar systems, the long-term thermal stability can be raised again to the values of unfilled materials; however, a higher stabilizer loading is compulsory.
Recently, it has been demonstrated that by using a polymeric glycidyl group containing copolymer at a loading ranging between 0.3–2 wt % as filler deactivator not only the thermal stability (OIT) could be improved from 30.3 to 96.6 min but even also the tensile impact strength value increased from 128 to 178 kJ/m2 for the polypropylene-based nanocomposite formulations.100
In general, polymer nanocomposites exhibit inferior light stability owing to the sorption of stabilizers or owing to metal ion-induced hydroperoxide decomposition. This can be partially circumvented by the use of UV absorbers and/or metal deactivators. Thus, UV absorbers of benzotriazole, benzophenone, or hydroxyphenyltriazine structures extended decisively the lifetime of linear low-density polyethylene (LLDPE) nanocomposite films. However, a metal deactivator alone outperformed the UV absorber in these experiments, indicating that the influence of metal impurities is also very crucial,90, 101 whereas the combined effect of both UVA and metal deactivator only marginally improved the photostability of the LLDPE nanocomposite. In addition to polyolefins,102 it has been experimentally demonstrated that polyamide nanocomposites can be synergistically stabilized by low- and high-molecular-weight light stabilizers in combination with processing and long-term heat stabilizers.
Alternatively, high light stability of nanocomposites can be reached with appropriate filler deactivators, that is the time until 50% retained tensile impact strength was raised from 1100 up to 2100 h in the presence of a filler deactivator. Artificial weathering of these stabilizer compositions reveals that the filler deactivator contributes by itself to the extended light stability and retention of mechanical properties.64, 71 Further improvement of light stability was possible by introducing an UV absorber.