FLAME RETARDANT DEVELOPMENTS
FOR POLYPROPYLENE

Edward A. Myszak, Jr. and Michael T. Sobus

Market Development Project Manager and Business Manager

PQ Corporation

Valley Forge, PA

Contact:

Nyacol Nano Technologies, Inc.

P.O. Box 349

Ashland, MA 01721

ABSTRACT

Flame retardants composed of solid inorganic particles generally have an adverse effect on the physical and aesthetic properties of polymers and synthetic fibers. This paper will discuss the advantages offered by colloidal-sized particles for flame retarding and UV stabilizing polypropylene materials. Particular attention will be given to the importance of choosing the whole flame retardant system, halogen and antimony oxide, and their affect on flame retardance, aesthetics and the polymer processing system.

INTRODUCTION

Advances in fine denier polypropylene fiber processing have opened the carpet face and wall covering markets to this versatile polymer. Aesthetically pleasing polypropylene structural products, such as lamp housings and other containers, have found a niche in the marketplace.

At the same time, today’s consumer is demanding flame retardant goods. Manufacturers need to supply products that are not only flame retarded but which also maintain their physical and aesthetic properties. With this in mind, we will divide our discussion into the flame retardant development hase and the UV development phase to simplify the variables introduced.

FLAME RETARDANT DEVELOPMENT

There currently are two systems that effectively flame retard polymers: halogenated systems and non-halogenated systems. Many manufacturers would prefer a non-halogenated system, such as magnesium hydroxide, aluminum trihydrate, ammonium phosphate, etc., principally because halogenated systems have received negative publicity. Non-halogenated systems, however, require loadings of up to 60% of flame retardant, and physical and aesthetic properties of the base polymers are negatively affected as a result. In the case of fine-denier fibers, a usable fiber could not be produced with these high flame retardant loadings. Halogenated systems offer the advantage of lower loadings due to the synergistic effect between the halogenated organic compound and antimony oxide, to achieve the desired levels of flame retardancy. In fact, several fiber manufacturers have dictated that no more than 8 % active FR ingredients can be used in a finished fine denier fiber. Most halogenated FR compounds and all non-halogenated systems, however, cannot meet this criteria.

An ideal flame retardant system would be easily processed and would safeguard the physical and aesthetic properties of a polymer. It would incorporate a melt-blendable halogenated additive with a sub-micron antimony oxide particle. This combination should yield an acceptably flame-retarded product with good tensile strength, impact resistance and elongation. The finished product should also be untinted / translucent.

An antimony pentoxide powder that disperses to colloidal size (0.03 micron) particles is the only FR additive that meets all these criteria. A detailed comparison of colloidal-sized antimony pentoxide versus antimony trioxide (the smallest particle size commercially available) is given in Table I.

TABLE I

Physical Properties of Antimony Pentoxide vs. Trioxide

Property	Antimony Trioxide	Antimony Pentoxide
Formula	Sb2O3	Sb2O5
Solubility	dilute acids & bases	only concentrated, hot acids
Particle Size	0.8 - 1.0 microns	0.03 microns
Surface Area m^2/gm	2	50
Specific Gravity	5.3	4.0
Refractive Index	2.1	1.7
Surface Activity	usually neutral	weakly acidic

Figure 1 shows the visual impact of a submicron antimony pentoxide article on a 1.5 denier polypropylene fiber vs. antimony trioxide. The antimony pentoxide particle occupies only 0.2% of the cross-sectional area of fiber vs. 7% for antimony trioxide.

Figure 1. - Antimony Pentoxide vs. Trioxide in a Fine Denier Fiber

RESULTS AND DISCUSSION

We tested the desirability of several potentially acceptable antimony oxide/halogenated additive systems for polypropylene fiber and translucent products. The antimony oxides used in this screening study were:

1. Nyacol^® ADP480 - a powder which disperses to colloidal size (0.03 micron) particles in nonpolar hydrocarbons

2. Antimony trioxide powder

Both antimony oxides were individually compounded into a flame retardant concentrate with each of five halogenated additives:

1. Brominated aromatic ester (63% Br)

2. Brominated polystyrene (60% Br)

3. Brominated polystyrene (66% Br)

4. Brominated aromatic compound (66% Br)

5. Chlorinated paraffin (74% Cl)

We produced concentrates of each combination to test the complete dispersion of the ingredients when the product was let down into the polymer. This is critical in fine-denier fiber applications. The concentrates contained 50% active FR ingredients in a carrier of polypropylene. This carrier was chosen to maintain the physical properties for fine denier fiber applications.

Table II summarizes the flame tests results of this initial screening of FR compounds. Only one halogen/antimony oxide compound exhibited flame retardancy according to the UL-94 vertical flame test. That one system is antimony pentoxide or trioxide with a brominated aromatic compound.

TABLE II

Flame Test Summary of Polypropylene

Additive Material	Percent^7 Additive	LOI^8 Percent	UL-94^9 Test	Afterflame Rating Time (Sec)
Virgin PP	NA	17.3	Fail	NA
BAE^1	12.0	22.7	Fail	NA
BP-60^2	12.0	19.3	Fail	NA
BAC^4	8.0^6	23.1	Fail	NA
ADP480^10/BAE	could not extrude
Trioxide BAE	12.0	22.9	Fail	NA
ADP480/BP-60	12.0	18.5	Fail	NA
Trioxide/BP-60	12.0	20.3	Fail	NA
ADP480/BP-66^3	12.0	18.5	Fail	NA
Trioxide/BP-66	12.0	20.5	Fail	NA
ADP480/CP^5	2.5	19.6	Fail	NA
BurnEx 2000 /BAC	12.0	23.6	V-2	0.0
ADP480/BAC	8.0	28.6	V-2	0.3
	4.0	26.9	V-2	9.4
	2.5	25.0	V-2	2.9
	1.0	20.1	V-2	3.8
Trioxide/BAC	12.0	32.3	V-0	0.0
	8.0	32.9	V-0	0.0
	4.0	28.9	V-2	0.0
	2.5	24.6	V-2	0.2
	1.0	21.6	Fail	NA
1 Brominated Aromatic Ester (63% Br)
2 Brominated Polystyrene (60% Br)
3 Brominated Polystyrene (66% Br)
4 Brominated Aromatic Compound (66% Br)
5 Chlorinated Paraffin (74% Br)
6 Could not extrude at 12% loading
7 Data for % Additive less than 12 not reported if sample failed UL-94
8 Limiting Oxygen Index (ASTM D2860) 9 Vertical Burn Test
10 ADP480 is a colloidal sized antimony pentoxide

The antimony pentoxide compounds were all rated V-2 according to UL-94 with afterflame times ranging from 0 to 3.8 seconds, depending on FR concentration. The trioxide compounds were rated V-O through FAIL depending on FR concentration. The V-O ratings achieved by trioxide at high loading levels (8 and 12%) were probably due to the rheology of the polymer being changed as a result of the trioxides’ larger particle size, which reduced the quantity of drips as well as their flaming characteristics.

A summary of the physical property test results of the UL-94 acceptable materials is given in Tables III and IV. The results with 1/8" thick test pieces show the antimony pentoxide and trioxide to be reasonably comparable from the perspectives of elongation and tensile strength. It would be expected that the larger trioxide particles would have a negative effect on these characteristics as the thickness of the test piece decreased.

TABLE III

Physical Property Summary

Additive Material	Percent Additive	Notched^2 Izod Impact (ft-lb/in)
Virgin PP	NA	0.64
BurnEx 2000	12.0	0.62
ADP480/BAC^1	8.0	0.63
	4.0	0.58
	2.5	0.64
	1.0	0.65
Trioxide/BAC	12.0	0.44
	8.0	0.36
	4.0	0.35
	2.5	0.37
	1.0	0.37
1 Brominated Aromatic Compound (66% Br)
2 ASTM D256

TABLE IV

Physical Property Summary

Additive Material	Percent Additive	Elongation^3 at Yield Percent	Tensile^3 Strength at Yield PSI	Elongation^2 at Break Percent	Tensile^2 Strength at Break PSI
Virgin PP	NA	16.8	4937	204.0	3055
BurnEx 2000	12.0	8.0	5168	29.4	3453
ADP480/BAC^1	8.0	9.9	5087	29.5	3643
	4.0	11.4	5173	30.3	3862
	2.5	15.9	5144	29.9	3564
	1.0	16.1	5068	49.6	3045
Trioxide/BAC	12.0	9.9	5040	32.5	3083
	8.0	11.5	4863	37.4	2841
	4.0	13.8	4892	31.9	2931
	2.5	13.7	5141	33.1	3265
	1.0	14.8	5122	30.9	3365
1 Brominated Aromatic Compound (66% Br)
2 ASTM states that tensile strength and elongation at break value for unreinforced polypropylene plastics generally are highly variable due to inconsistencies in necking of the center section of the test bar. Tensile strength and elongation at yield are more reproducible.
3 ASTM D638

The Izod impact data, however, show that the material with pentoxide has a significant advantage at all loading levels. In fact, the Izod data for polypropylene processed with antimony pentoxide-based flame retardants are comparable to the Izod result for virgin PP.

Figure 2 shows the advantage of the smaller antimony pentoxide particles. The Izod data is 38 to 75% better for the Nyacol ADP480 compounds than for the trioxide-based compounds. As with elongation and tensile strength data, we strongly believe that, as the test piece thickness decreases, the difference between pentoxide- and trioxide- based compounds will become even more exaggerated in favor of pentoxide-based compounds.

The antimony pentoxide- and trioxide-based flame retardant compounds processed equally well at all loading levels except at 12%, where the trioxide-based flame retardant compounds processed more easily. The larger trioxide particles may have absorbed the halogen material and thereby prevented puddling or slippage in the throat of the extruder. Ease of processing is probably a moot issue, however, since the industry standards for FR loading levels are expected to be no higher than 8%.

Figure 2. - Izod Results

Table V shows the color effects of the flame retardant additive on the polymer. This data is reported as L'a'b' total color difference as measured on a Minolta CR-200 Chroma Meter with virgin polymer as the base standard. The color is reported in the standard CIE 1976 L'a'b' notation.

TABLE V

Color Effect of Additive on Polymer - Unpigmented

Additive Material	Percent Additive	L'a'b' Total Color Difference	Translucency (Visual)
Virgin PP	NA	0.0	Translucent
BAC	4.0	41.2	Opaque
ADP480	4.0	34.2	Slightly Translucent
Trioxide	4.0	58.7	Opaque
BurnEx 2000	12.0	51.1	Opaque
ADP480/BAC	8.0	50.1	Opaque
	4.0	43.7	V Slightly Translucent
	2.5	37.2	Slightly Translucent
	1.0	19.4	Translucent
Trioxide/BAC	12.0	54.4	Opaque
	8.0	53.4	Opaque
	4.0	52.1	Opaque
	2.5	47.1	Opaque
	1.0	35.7	V Slightly Translucent

Figure 3 graphically shows a dramatic difference between Nyacol ADP480-based compounds and those processed with trioxide. ADP480 has less pigmenting or whitening effect than trioxide on the base polymer.

Figure 3. - Color Effect of FR Addition on Virgin PP - Unpigmented

A comparison of antimony pentoxide versus trioxide reveals a color difference at 4% loading of 34.2 versus 58.7, respectively. At 4% loading the FR additive system utilizing antimony pentoxide shows the test piece at 1/8 inch thickness begins to exhibit some translucency (L'a'b' delta of 43.7). The level of translucency for the antimony pentoxide compounds increases as the loading level drops to 1% (L'a'b' delta of 19.4). The antimony trioxide compounds are opaque at all loading levels except at a 1% loading, where the test pieces exhibit only slight translucency (L'a'b' delta of 35.7), but the material fails UL-94. 

One of the curiosities of physics, which we will not attempt to explain, is that very small and very large particles have low hiding power, or opacity. There is, on the other hand, an optimum size for maximum opacity. For antimony trioxide, the 0.5 - 1 micron particles provide maximum opacity. Figure 4 dramatizes the translucence at 4% loadings of antimony pentoxide vs. antimony trioxide when compounded with virgin polypropylene. The affects of a 2.5% FR compound pentoxide vs. trioxide are also shown.

Figure 4. - Translucency Effect of Flame Retardant 
Addition to Virgin PP

Differences in opacity levels in unpigmented environments help explain why less color concentrate is required to obtain a given color in an antimony pentoxide vs. trioxide flame retarded material. Generally a polymer system that uses antimony trioxide as a flame retardant will require an average of four times more pigment to achieve a particular shade than a polymer system that contains antimony pentoxide. Even higher proportions of pigment are needed to achieve dark red and blue tones when antimony trioxide is involved (see Figure 5). Dollar savings for pigments can be significant when antimony pentoxide is used for flame retardancy.

Figure 5. Color Difference - Antimony Pentoxide vs. Trioxide

With the "ideal" flame retardant developed, the next area to consider is the UV stability of the system and, if required, how to improve it while not affecting or minimizing the affects on the flame retarded end product.

UV STABILITY DEVELOPMENT

All of us have had contact with UV absorbers in our daily lives, i.e., sunscreens, sunglasses and a host of other retail products. We understand that UV radiation from the sun causes our skin to tan and even burn. In all likelihood, however, we do not think about the effects of UV radiation on such things as automotive paints, plastic outdoor furniture, and, least of all, flame retarded fabrics or parts.

UV radiation is electromagnetic radiation in the part of the spectrum between x-rays and visible light. Its wavelengths are longer than those of x-rays and shorter than those of ordinary light. It differs from light only in that its wavelengths are too short to be visible to the human eye. It is readily seen by certain insects, i.e., bees.

The UV spectrum is divided into the near, the far, and the vacuum UV regions. These regions include, respectively, the wavelengths from 4,000 to 3,000Å (UVA), 3,000 to 2,000Å (UVB), and 2,000 to 40Å (UVC). The region between 4,000 to 3,000Å is sometimes referred to as the region of "black light."

The chief natural source of UV radiation is the sun. About 9% of all the energy emitted by the sun is UV radiation. Most of this is in the region between 4,000 and 3,000Å, with only about 14% consisting of wavelengths shorter than 3000Å. More than half of the solar UV radiation that reaches the earth is absorbed by the earth’s atmosphere, namely, the ozone layer. This absorption is greater for the shorter than for the longer wavelengths, so that almost none of the solar energy with wavelengths shorter than 3000Å reaches the surface of the earth.

Like visible light, UV radiation obeys the laws of optical reflection and refraction. However, it is absorbed by many substances that are transparent to visible light. Ordinary window glass, for example, is almost opaque to wavelengths shorter than 3000Å. Likewise, most clean, thin-layered plastics will transmit some UV radiation in the region between 4,000 and 3,000Å, but they are almost opaque to shorter wavelengths.

UV radiation promotes many chemical reactions, the study of which constitutes part of the field of photochemistry. The fading and bleaching of dyes exposed to sunlight are instances of these reactions. Another example, which is of prime importance for this discussion today, is the reaction of a halogenated (in particular, bromine) flame retardant system when exposed to UV radiation. In general, we expect the subject material to yellow when exposed to UV radiation. The higher the radiation dose, the more yellow the material will become. The effects are cumulative.

There is an additional negative synergistic effect when a flame-retarded material is abused thermally, i.e., when processed at excessive temperatures. When that thermally abused material is exposed to UV radiation, the yellowing effect is exaggerated. The higher the process temperature without the aid of an antioxidant, the faster and more intense the yellowing of the material will be when finally exposed to UV radiation.

Therefore, the part or fiber must be protected from UV-induced degradation. To maintain the light fastness of the colored material, we developed a submicron UV absorbing product, namely, zinc oxide. Zinc oxide has been used for many years in sunscreen formulations as a UV absorber to protect the skin. It has also been used in paint formulations. However, as you all remember, zinc oxide behaves as a white pigment and results in opaque final products, i.e., white-nosed lifeguards and opaque "rough" (flat) paint finishes. To obviate this drawback, we have developed a 20-50nm particle size zinc oxide in aqueous systems (DP5370) and a 110nm particle size zinc oxide in a dry powder system (DP5372). We will concentrate on the dry powder system since we plan to use this absorber in an extruded polymeric application.

Table VI details the DP5372 loadings we tested in conjunction with the flame retardant of choice, Nyacol BurnEx 2000, to achieve effective UV absorption.  

TABLE VI

Loadings Tested with Nyacol^® BurnEx 2000

Virgin PP	BurnEx 2000^1	DP5372^2
100%	-	-
98%	2%	-
98%	2%	2%
98%	2%	5%
98%	2%	10%
98%	2%	0.5%