Flame Retardant Polypropylenes: A Review

Polypropylene (PP) is a commodity plastic known for high rigidity and crystallinity, which is suitable for a wide range of applications. However, high flammability of PP has always been noticed by users as a constraint; therefore, a variety of additives has been examined to make PP flame-retardant. In this work, research papers on the flame retardancy of PP have been comprehensively reviewed, classified in terms of flame retardancy, and evaluated based on the universal dimensionless criterion of Flame Retardancy Index (FRI). The classification of additives of well-known families, i.e., phosphorus-based, nitrogen-based, mineral, carbon-based, bio-based, and hybrid flame retardants composed of two or more additives, was reflected in FRI mirror calculated from cone calorimetry data, whatever heat flux and sample thickness in a given series of samples. PP composites were categorized in terms of flame retardancy performance as Poor, Good, or Excellent cases. It also attempted to correlate other criteria like UL-94 and limiting oxygen index (LOI) with FRI values, giving a broad view of flame retardancy performance of PP composites. The collected data and the conclusions presented in this survey should help researchers working in the field to select the best additives among possibilities for making the PP sufficiently flame-retardant for advanced applications.


Introduction
Polymers are building blocks of advanced materials and systems, but their flammability has been a serious constraint in their usage in advanced applications [1][2][3]. Polypropylene (PP) is a commodity plastic widely used in a variety of applications, particularly in the form of composites in load-bearing uses due to its high rigidity and crystallinity [4]. By the end of 2020, the PP market size is expected to reach $112 billion, and it is estimated to reach $155 billion by 2026 [5,6]. Its global production was 56.0 million metric tons in 2018, and it is estimated to reach around 88.0 million metric tons by 2026. This growing demand reflects the importance of PP for applications where low density, hardness, high flexural modulus, and chemical resistance are needed [7,8] (1) Basically, the use of FRI makes it possible to semi-qualitatively classify polymer composites by labeling them as Poor, Good, or Excellent flame retardancy performance and thus enables evaluation of the efficiency of the incorporated flame retardant (FR). There has always been a need for fast-tracking and classifying polymers for their flame retardant performance. The use of FRI made possible classifying polymers and polymer composites in terms of flame retardancy in a simple manner. For FRI values below 10 0 obtained by the use of Equation (1), we have the case (namely Poor) where the addition of FR adversely affects flame retardancy of polymer. When FRI takes values in the range of 10 0 -10 1 , we name it Good flame retardancy performance, such that addition of FR enhances the resistance of polymer against fire. For FRI values above 10 1 , which is rare in practical cases, we have an Excellent case, where FR significantly improves flame retardancy. It is worth mentioning that some important parameters of testing such as irradiance and sample thickness as well as sample weight can be neglected due to the fact that, in the FRI formula, the parameters related to the neat polymer are divided by those of polymer/FR composite. Thus, the dimensionless value obtained can be used as a reliable measure of the efficiency of FR in polymer. In this survey, the data from the literature were extracted first, and five families of flame retardants that served as PP were considered including phosphorus-based, nitrogen-based, mineral, carbon-based, and bio-based flame retardants, and hybrid cases composed of the aforementioned five categories were distinguished. The main aim of the present survey is to give the readers a broad view of FR systems used in PP via FRI classification method. Certainly, this classification is not a precise and unique data set for FR selection for PP, but it can be considered as a database to compare different systems. The focus of this work was particularly placed on the reports in which cone calorimetry test was carried out. However, some other parameters such as smoke quantity or the percentage of FR elements (phosphorus, nitrogen, . . . ) were not systematically given in this research paper due to the lack of data, which could lead to unreliable judgments. For some papers, limiting oxygen index (LOI) and UL-94 data were also available, which were used in finding possible correlations between the FRI variation and other criteria.

Phosphorus-Based Flame Retardants
Various types of phosphorus-based flame retardants have been incorporated into PP to make it flame-retardant [21][22][23]. Table 1 reviews the names and the percentages of these flame retardants Table 1. Flame-retardant PP materials containing phosphorus-based (P) flame retardants. Data are extracted from the literature: cone calorimetry parameters (TTI, pHRR, THR), LOI, and UL-94 values.

Nitrogen-Based Flame Retardants
Nitrogen-based flame retardants have also been used in PP to make it resistant against fire. Table 2 gives the names and the percentages of incorporation of these flame retardants, where the data were obtained in cone calorimetry (pHRR, THR, and TTI), FRI calculated from cone calorimetry parameters, as well as LOI and UL-94 values. Some of the nitrogen-based FRs listed in Table 2 also contain a Polymers 2020, 12, 1701 7 of 49 phosphorus element. However, the percentage of nitrogen is more important, and therefore these FRs are listed in this Table. The FRI values were calculated by authors of the present review. The name and the percentage of flame retardants are provided in separate columns. "wt.%" was used for loading level of additives, while "-" stands for the systems free of additive or the neat PP. * FR means flame retardant. Since all comparisons were made in terms of FRI, classification of polymers in terms of their flame-retardant properties was not surveyed based on the chemistry of additives, heat flux, sample thickness, etc.  To give a bright view of the variation trend, Figure 4 illustrates the FRI values as a function of wt.% of nitrogen-based flame retardants incorporated into the PP. The percentage of incorporation was changed from 15 to 40 wt.%. Of note, all points are located in the Good zone of FRI, except two points remarked as Excellent. These two points correspond to a kaolinite additive modified with nitrogen and phosphorus agents. A very noticeable point to be considered is that increasing the amount of diallyldimethylammonium (nominated with the symbol in Figure 4) from 5 to 25 has no serious effect on the value of FRI, so that they are aligned vertically around FRI values between 1.0 and 2.5. Overall, , TA-CFA-30 [39], TA-CFA-30 [40], TA-CFA-25 [37], TA-CFA-25 [32], TA-CA-ZnO-25 [29], TA-CA-25 [67], TA-CA-20 [27], TA-CA-20 [23], TA-IFR-20 [68], TA-IFR-25 [31], PI-FR-25 [69], PI-IFR-30 [70], PI-IFR-20, PI-IFR-30, PI-IFR-40 [71], NOR116-0.5 [72], NOR116-0.3 [73], PPU-CA-25 [50], N-FR-22, N-FR-25 [74], N-IFR-5, N-IFR-10, N-IFR-15, N-IFR-20, N-IFR-25 [75], N-IFR-25 [76], PN-IFR-30 [56]. Figure 5 patterns UL-94 results as a function of FRI for nitrogen-based flame retardant in PP. It can be observed that even at small quantities of FRI, V0 in UL-94 was achieved. The diversity of data in Figure 5 can be taken as a signature of sensitivity of UL-94 to FRI. Figure 6 shows LOI values as a function of FRI. There is a quite reasonable correlation between the LOI and FRI values, up to FRI value of 6.

Mineral-Based Flame Retardants
Mineral additives have been widely used in polymers for their acceptable cost and properties [79]. Mineral-based flame retardants including clays are widely used in PP due to their low cost and acceptable thermal resistance. In this family, the most used flame retardants in volume were aluminum trihydroxide (ATH) and magnesium dihydroxide (MDH). However, due to their low efficiency, a high percentage of loading was necessary for achieving an acceptable level of flame retardancy of polymers. The name and the percentage of the used mineral-based flame retardants in PP are listed in Table 3. Cone calorimetry data, FRI, LOI, and UL-94 values are also given so as to make possible a detailed view on the status of flame retardant efficiency of PP materials. The FRI values were calculated by authors of the present review. The name and the percentage of flame retardants are provided in separate columns. "wt.%" was used for loading level of additives, while "-" stands for the systems free of additive or the neat PP. * FR means flame retardant. Since all comparisons were made in terms of FRI, classification of polymers in terms of their flame-retardant properties was not surveyed based on the chemistry of additives, heat flux, sample thickness, etc.      Figure 7 visualizes the variation of FRI value as a function of flame retardant loading in PP systems (for the convenience of readers, two figures are added for giving a better zoom on data points). This figure clearly shows that even at low loading percentages, it is possible to achieve a relatively high FRI value depending on the type of mineral. There is no denying that some parameters such as the state of dispersion and size of particles are important factors affecting the flame retardant properties.   Unfortunately, the number of papers in which cone calorimetry, UL-94, and LOI values were studied was indeed limited, but the ones available are used plotting Figure 8. It should be noted that no formulation among studied ones is rated at V0. In conclusion, it is quite difficult to find a correlation between quantitative and qualitative parameters based on such a tiny set of data. In regard to the relationship between LOI and FRI, a meaningful trend can still be seen in Figure 9.

Carbon-Based Flame Retardants
Carbon-based additives have been widely used in developing polymer composites and nanocomposites [118][119][120][121]. However, due to expense and limited interaction with PP, a few works based on carbon-based flame retardants have been reported on flame-retardant PP materials. Table 4 summarizes all information available on the flame-retardant PP materials containing carbon-based additives. Table 4. Flame-retardant PP materials containing carbon-based (C) flame retardants. Data are extracted from the literature: cone calorimetry parameters (TTI, pHRR, THR), LOI, and UL-94 values. The FRI values were calculated by authors of the present review. The name and the percentage of flame retardants are provided in separate columns. "wt.%" was used for loading level of additives, while "-" stands for the systems free of additive or the neat PP. * FR means flame retardant. Since all comparisons were made in terms of FRI, classification of polymers in terms of their flame-retardant properties was not surveyed based on the chemistry of additives, heat flux, sample thickness, etc.   Figure 10 shows that with low loading percentage (1 wt.%) of carbon nanotubes, it is possible to achieve the Good FRI. No data were available for UL-94 tests. Comparison between Figures 7 and 10 also suggests that low-cost minerals were used at higher loadings, while carbon-based additives were used almost at loadings below 10 wt.%. A limited number of data have also been reported on LOI values. These points are plotted as a function of FRI in Figure 11, where a good correlation can be established between FRI and LOI values. Deeper understanding of the mechanism behind such correlation requires a detailed view of the origin of tests as well as the chemical structure of additives and possible interaction between the PP and additives.

Bio-Based Flame Retardants
In recent years, due to sustainability issues, the use of bio-based additives has also been investigated in PP. However, the number of research papers is limited on this subject. Table 5 gives the name and loading percentage of these bio-based FR. The obtained results from cone calorimetry, LOI, and UL-94 tests are also listed in Table 5. Figures 12 and 13 display UL-94 and LOI results as a function of FRI for bio-based flame retardant in PP, respectively.     FRI values are plotted as a function of loading percentage of bio-based FR in Figure 14. It can be observed that a high quantity of bio-based FR, 40 wt.% is needed to achieve FRI equal to 6.

Combination of Flame Retardants
As observed in previous sections, using an additive alone can to a limited extent improve flame-retardant properties of PP. Combination of flame retardants is a strategy to improve further the flame retardancy via synergism between various flame retardants [140][141][142]. Moreover, the quantity of the used flame retardant can be reduced in polymer so as to prevent mechanical properties deterioration. Different combinative additive systems were considered in PP. The corresponded data are collected and summarized in Table 6. The third column gives the ratio between flame retardants. Table 6. The flame retardancy performance of PP containing various combinations of flame retardants in terms of FRI (* the name and percentage of incorporated flame retardants are given after PP). P = phosphorus FR, Np = non-phosphorus FR, N = nitrogen FR, nN = non-nitrogen-based FR, M = mineral FR, Bio = bio-based FR, nBio = non bio-based FR (one can also consider some nitrogen-based FRs containing phosphorus element as the combination of phosphorus and nitrogen resulting in synergism, Table 2). Since all comparisons were made in terms of FRI, classification of polymers in terms of their flame-retardant properties was not surveyed based on the chemistry of additives, heat flux, sample thickness, etc.                Figure 15 displays the performance of different combinatorial additive systems used for PP. It can be clearly observed from the left-hand side figure that cases with FRI values above 10 (Excellent zone) are more frequent compared to all previous cases in which only one additive was used. More interestingly, the combination of additives appeared a useful strategy where very high FRI values (event more than 50) took place at intermediate loadings (25)(26)(27)(28)(29)(30) wt.%). For achieving a high FRI value, the combination of several types of flame retardants is needed, for example, phosphorus, intumescent, and mineral flame retardants [150] or phosphorus, nitrogen, and mineral flame retardants [164]. Figure 16 shows that V-0 level in UL-94 is automatically obtained in the case of combined flame retardant systems used in PP regardless of the FRI value. However, no correlation exists between the FRI and LOI (Figure 17). The complexity of polymer-filler interaction can be considered as the main reason for diversity of properties. zone) are more frequent compared to all previous cases in which only one additive was used. More interestingly, the combination of additives appeared a useful strategy where very high FRI values (event more than 50) took place at intermediate loadings (25)(26)(27)(28)(29)(30) wt.%). For achieving a high FRI value, the combination of several types of flame retardants is needed, for example, phosphorus, intumescent, and mineral flame retardants [150] or phosphorus, nitrogen, and mineral flame retardants [164]. zone) are more frequent compared to all previous cases in which only one additive was used. More interestingly, the combination of additives appeared a useful strategy where very high FRI values (event more than 50) took place at intermediate loadings (25)(26)(27)(28)(29)(30) wt.%). For achieving a high FRI value, the combination of several types of flame retardants is needed, for example, phosphorus, intumescent, and mineral flame retardants [150] or phosphorus, nitrogen, and mineral flame retardants [164]. zone) are more frequent compared to all previous cases in which only one additive was used. More interestingly, the combination of additives appeared a useful strategy where very high FRI values (event more than 50) took place at intermediate loadings (25)(26)(27)(28)(29)(30) wt.%). For achieving a high FRI value, the combination of several types of flame retardants is needed, for example, phosphorus, intumescent, and mineral flame retardants [150] or phosphorus, nitrogen, and mineral flame retardants [164]. retardancy via synergism between various flame retardants [140][141][142]. used flame retardant can be reduced in polymer so as to prevent mechan Different combinative additive systems were considered in PP. The cor and summarized in Table 6. The third column gives the ratio between f Table 6. The flame retardancy performance of PP containing various comb in terms of FRI (* the name and percentage of incorporated flame retarda phosphorus FR, Np = non-phosphorus FR, N = nitrogen FR, nN = non-nitro FR, Bio = bio-based FR, nBio = non bio-based FR (one can also conside containing phosphorus element as the combination of phosphorus synergism, Table 2). Since all comparisons were made in terms of FRI, c terms of their flame-retardant properties was not surveyed based on the c flux, sample thickness, etc. Figure 15 displays the performance of different combinatorial add can be clearly observed from the left-hand side figure that cases with F zone) are more frequent compared to all previous cases in which only interestingly, the combination of additives appeared a useful strategy (event more than 50) took place at intermediate loadings (25)(26)(27)(28)(29)(30) wt.%). F the combination of several types of flame retardants is needed intumescent, and mineral flame retardants [150] or phosphorus, n retardants [164]. retardancy via synergism between various flame retardants [140][141][142]. Moreover, the quantity of the used flame retardant can be reduced in polymer so as to prevent mechanical properties deterioration. Different combinative additive systems were considered in PP. The corresponded data are collected and summarized in Table 6. The third column gives the ratio between flame retardants. Table 6. The flame retardancy performance of PP containing various combinations of flame retardants in terms of FRI (* the name and percentage of incorporated flame retardants are given after PP). P = phosphorus FR, Np = non-phosphorus FR, N = nitrogen FR, nN = non-nitrogen-based FR, M = mineral FR, Bio = bio-based FR, nBio = non bio-based FR (one can also consider some nitrogen-based FRs containing phosphorus element as the combination of phosphorus and nitrogen resulting in synergism, Table 2). Since all comparisons were made in terms of FRI, classification of polymers in terms of their flame-retardant properties was not surveyed based on the chemistry of additives, heat flux, sample thickness, etc. Figure 15 displays the performance of different combinatorial additive systems used for PP. It can be clearly observed from the left-hand side figure that cases with FRI values above 10 (Excellent zone) are more frequent compared to all previous cases in which only one additive was used. More interestingly, the combination of additives appeared a useful strategy where very high FRI values (event more than 50) took place at intermediate loadings (25)(26)(27)(28)(29)(30) wt.%). For achieving a high FRI value, the combination of several types of flame retardants is needed, for example, phosphorus, intumescent, and mineral flame retardants [150] or phosphorus, nitrogen, and mineral flame retardants [164]. perties of PP. Combination of flame retardants is a strategy to improve further the flame ia synergism between various flame retardants [140][141][142]. Moreover, the quantity of the etardant can be reduced in polymer so as to prevent mechanical properties deterioration.
binative additive systems were considered in PP. The corresponded data are collected ized in Table 6. The third column gives the ratio between flame retardants.
The flame retardancy performance of PP containing various combinations of flame retardants of FRI (* the name and percentage of incorporated flame retardants are given after PP). P = rus FR, Np = non-phosphorus FR, N = nitrogen FR, nN = non-nitrogen-based FR, M = mineral = bio-based FR, nBio = non bio-based FR (one can also consider some nitrogen-based FRs g phosphorus element as the combination of phosphorus and nitrogen resulting in m, Table 2). Since all comparisons were made in terms of FRI, classification of polymers in their flame-retardant properties was not surveyed based on the chemistry of additives, heat ple thickness, etc.  The vertical intervals in each category, i.e., V-0, V-1, V-2, and NR, are schematically representative of the amount of additive used. For example, two data distinguished by different symbols having the same or very close FRI values (horizontal quantity) in a given category (e.g., V-1), may have different vertical quantities, e.g., both reveal V-1 behavior in UL-94 test, but the upper contains more FR in PP.   The left-side plot reveals that FRI values above 10 (Excellent zone) took place in several cases, which is in contradiction with all previous cases in which only one additive was used.

Conclusions and Future Perspective
This work opens new avenues to the experts working on "flame retardant polyolefins", the title of a Special Issue entitled "Flame Retardant Polyolefins" in Polymers journal for which this work is designed and carried out. In this work, more than 150 research papers from the literature dealing with the flame retardancy of PP were analyzed, classified, and discussed in terms of flame retardancy performance. From the selected papers were extracted cone calorimetry data to calculate Flame Retardancy Index (FRI) as a measure or label of flame retardant performance. To have a comprehensive overview of flame retardant PP materials, works on PP flame retardancy were categorized in terms of additives used in classes including: phosphorus-based, nitrogen-based, mineral, carbon-based, bio-based, and hybrid combinatorial flame retardants composed of two or more additives. The analysis of efficiency of flame retardancy was performed in terms of the FRI variation as a function of wt.% of additives used. The analysis quite obviously unveiled the superiority of the combination of additives over the use of each one separately. In addition, the UL-94 and LOI values available in each class of additives were plotted in terms of the FRI so as to find possible correlation between analyses made in the literature. This work provided a pool of data on flame-retardant PP materials for future research on PP materials. It was elucidated that FRI can satisfactorily make possible classification of PP materials in terms of flame retardancy performance. The present work provides those research works that claim achieving synergistic effect of two or more flame retardants with a clear measure of flame retardant performance as Poor, Good, and Excellent labels assigned to PP materials, based on cone calorimetry data. Moreover, future works on LOI and UL-94 tests can be added to the data used here so as to draw a more detailed picture of flame retardancy behavior of PP materials. The approach can be used to make judgement about other flame retardant polyolefins. Moreover, we believe that the mechanical properties of FR polymers should also be considered in the future, but it is pertinent to the completeness of data in the literature. The importance of mechanical properties springs from the fact that highly loaded systems are prone to mechanical failure as a consequence of stress concentration. All in all, the type and the percentage of FRs in polymers affect both the mechanical and flame retardant properties of polymers; therefore, optimization of both properties is of importance.