Alternative Aircraft Anti-Icing Formulations with Reduced Aquatic Toxicity and Biochemical Oxygen Demand (2010)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

4-1 SECTION 4 TIER 1 RESULTS The candidate aircraft deicing/anti-icing and airfield deicing formulation components identified for Tier 1 testing included 27 FPDs (5 aircraft, 10 runway, and 12 for both aircraft and runway), 5 thickeners, 20 surfactants, and 17 corrosion inhibitors. Quantities of each component sufficient for Tier 1 testing were ordered from various sources, including chemical distributors (Alfa Aesar, Chemical Marketing Concepts and Sigma-Aldrich) and chemical manufacturers (Air Products, BASF, CP Kelco, Dow, Lubrizol, PMC Specialties, and Stepan). Table 4-1 lists the numbers of components that were ordered, received, and available for testing. TABLE 4-1. Number of alternative deicing components ordered and available for testing. Component Ordered Not Available Available for Testing FPDs 27 1 26 Thickeners 5 0 5 Surfactants 20 1 19 Corrosion Inhibitors 17 3 14 The samples were divided into quantities suitable for aquatic toxicity, BOD and COD profiles, freezing points, flash point, surface tension and viscosity testing. Samples for aquatic toxicity testing, BOD, and COD were sent to the Wisconsin State Laboratory of Hygiene (Madison, WI); flash point samples were sent to Thorstensen Laboratories (Westford, MA); and the remainder to Infoscitex (Waltham, MA) and the University of Massachusetts -Lowell (Lowell, MA) for freezing point, surface tension and viscosity tests. Freezing Point Depression The following criteria were used in evaluating the freezing point depression performance for the mixtures of FPDs and water: • FPD mixtures frozen at -14.5°C were rejected from further evaluation • FPD mixtures not frozen at -14.5°C, but frozen at -20°C were evaluated for runway deicing applications • FPD mixtures not frozen at -20°C were evaluated for aircraft de-icing applications A simple freeze test was used to determine whether the FPD:water mixtures met the criteria. The freezing point depression of the candidate FPDs was obtained by dissolving each chemical in a 1:1 ratio with water. Four replicates of each solution were placed in small test tubes mounted in a rack and surrounded by dry ice in a vessel (Figure 4-1). The dry ice was doused with 2-propanol to lower its freezing point.

ALTERNATIVE AIRCRAFT ANTI-ICING FORMULATIONS 4-2 Figure 4-1. Rack apparatus for freezing point depression tests. The freezing point of each solution was measured using an alcohol thermometer. First, the solution was allowed to completely freeze. The rack was then removed from the vessel and the melting point for each solution was monitored and recorded. This was repeated for all samples analyzed. Table 4-2 shows the placement of the test tubes in the rack and the observed results for four FPDs. Measurements were made to determine whether the solutions met the criteria for freezing at -14.5°C and -20°C and the temperatures when they completely froze. Table 4-3 summarizes the results of the freezing point depression tests for all of the FPDs. TABLE 4-2. Typical results for freezing point depression. Freeze at °C? Test Tube Placement (See Location in Figure 4-1) -14.5°C -20.0°C Completely Freezes (°C) 1,3-Propylene Glycol: CAS No. 504-63-2 1 No No -50.0 to -52.0 14 No No -48 to -48.5 7 No No -50 19 No No -50 2,3- Butanediol: CAS No. 513-85-9 5 Yes Yes -6.0 12 Yes Yes -6.5 13 Yes Yes -6.5 20 Yes Yes -6.0 2-(2-Methoxyethoxy)-ethanol: CAS No. 111-77-3 16 Yes Yes -55.0

SECTION 4—3BTIER 1 RESULTS 4-3 TABLE 4-2. Typical results for freezing point depression. Freeze at °C? Test Tube Placement (See Location in Figure 4-1) -14.5°C -20.0°C Completely Freezes (°C) 9 Yes Yes -57.0 10 Yes Yes -55.0 8 Yes Yes -56.0 D-Gluconic acid, δ-lactone: CAS No. 90-80-2 15 Yes Yes -22.0 17 Yes Yes -22.0 6 Yes Yes -22.0 11 Yes Yes -24.0

ALTERNATIVE AIRCRAFT ANTI-ICING FORMULATIONS 4-4 TABLE 4-3. Results of freezing point depression and flash point testing for FPDs. FPD CAS Number Flash Point (°C) Freeze at °C State Aircraft or Runway Note 50:50 Neat -14.5°C -20°C Unfreeze Frozen at -14°C—Rejected for Further Evaluation Ethylene carbonate 96-49-1 >150 160 Yes Yes Freezes at 0°C Solid A, R Failed freezing point Propylene carbonate 108-32-7 135 132 Yes, in slush form Yes, in slush form Freezes completely at -30°C and slush forms from -4°C to - 30°C Liquid A Failed freezing point 2,3-Butanediol 513-85-9 95 85 Yes Yes -4°C Liquid A Failed freeze, BOD/COD 4-Methyl-γ- butyrolactone 108-29-2 >150 96 Yes Yes Freezes in slush from -4°C to -20°C Liquid A Failed freeze point Dimethyl malonate 108-59-8 90 90 Yes Yes Freezes as a white solid Liquid A Not miscible; failed freeze Dimethyl succinate 106-65-0 85 90 Yes Yes Frozen solid at 0°C Liquid A Failed; freeze point and not miscible Sodium acetate 127-09-3 >150 >249 Yes Yes Frozen completely at -4°C; freezes at -20°C for 2:1 and 3:1 water:sodium ratios; not frozen at -20°C for 4:1 and 5:1 water:sodium ratios; not frozen at -14.5°C for 2:1 to 5:1 water:sodium ratios at Solid R Failed freezing point at 1:1 (water:sodium) mixture; did not freeze at higher ratios Calcium propionate 4075-81-4 >150 High Yes Yes -5°C; completely frozen at -6°C Solid R Failed freezing point Disodium succinate 150-90-3 >150 High Yes Yes -5°C; Completely frozen at -6°C Solid R Failed freezing point

SECTION 4—3BTIER 1 RESULTS 4-5 TABLE 4-3. Results of freezing point depression and flash point testing for FPDs. FPD CAS Number Flash Point (°C) Freeze at °C State Aircraft or Runway Note 50:50 Neat -14.5°C -20°C Unfreeze Not Frozen at -14.5°C but Frozen at -20°C—Further Evaluated for Runway Deicing Applications 2,2-Dimethyl-1,3- dioxolane-4-methanol 100-79-8 >150 80 No Yes -24°C to -19°C; freezes in slush form at -20°C to -25°C Liquid A, R Failed freeze point for aircraft deicer; passed freeze point for runway deicer Xylitol 87-99-0 >150 High No Yes -21°C; mixes well with water; freezes as a white solid at -22°C; frozen is slush form at -20°C Solid R Passes freezing point test for runway deicer Not Frozen at -20°C—Further Evaluated for Aircraft and Runway Applications 1,2-Propylene glycol 57-55-6 100 107 No No Freezes at -65°C Liquid A, R — 1,3-Propylene glycol 504-63-2 98 >110 No No Freezes at -50°C Liquid A, R — Glycerol 56-81-5 >150 160 No No -25°C; freezes as a white solid at -50°C Liquid A, R — 1,3-Butanediol 107-88-0 110 121 No No -29°C to -26°C Liquid A, R — 2-Methyl-1,3- propanediol 2163-42-0 107 >110 No No -4.5°C Liquid A, R — Diethylene glycol 111-46-6 >150 143 No No Freezes as a white solid at -50°C; freezes in slush form at -20°C Liquid A. R — 2-(2-Methoxyethoxy)- ethanol 111-77-3 87 <103 >93 84 No No -50°C Liquid A, R. — 2-(2-Ethoxyethoxy)- ethanol 111-90-0 84 <103 >93 96 No No Not frozen at -50°C Liquid A, R —

ALTERNATIVE AIRCRAFT ANTI-ICING FORMULATIONS 4-6 TABLE 4-3. Results of freezing point depression and flash point testing for FPDs. FPD CAS Number Flash Point (°C) Freeze at °C State Aircraft or Runway Note 50:50 Neat -14.5°C -20°C Unfreeze Dipropylene glycol 25265-71- 8 139 135 No No Did not freeze at -60°C Liquid A, R — Triethylene glycol 112-27-6 180 165 No No -24°C Liquid A, R — 1,1,1- Trimethanolethane 77-85-0 148 160 — — Did not mix well with water; crystals do not blend completely with water Solid R Does not mix well with water D-Gluconic acid,δ- lactone 90-80-2 >150 High No No Solid to slush at -22°C to -21°C; slush to liquid at -15°C to - 8°C Solid R — Trimethylolpropane 77-99-6 >150 172 No No Mixes well with water; freezes completely at -22°C; slush forms between -12°C to -21°C Solid R — L-Tartaric acid dipotassium salt 921-53-9 >150 High No No Mixes well with water; frozen completely as a solid at -35°C; frozen in slush form between -34°C to -16°C Solid R — Tripotassium citrate 6100-05-6 >150 High No No Completely frozen at -45°C and in slush from at -42°C Solid R —

SECTION 4—3BTIER 1 RESULTS 4-7 The costs of the FPDs (on a neat basis) that were further evaluated in Tier 2 are shown in Table 4-4. The costs are based on the Sigma-Aldrich catalog as of August 5, 2008. The costs of large quantities of these chemicals were not readily available. The costs of commercially available current-use aircraft (1.2-propylene glycol) and runway (potassium acetate) deicers are included for comparison. TABLE 4-4. Cost of FPD on a neat basis (in order of increasing cost).a Application FPD Cost, $/kg Aircraft Triethylene glycol 14.80 1,3-Butanediol 17.00 2-Methyl-1,3-propanediol 18.90 Trimethylolpropane 23.87 Dipropylene glycol 26.00 Glycerol 33.30 Diethylene glycol 123.20 1,3-Propylene glycol 154.00 1,2-Propylene glycol 16.15 Runway Trimethylopropane 23.90 Sodium acetate 27.10 D-Gluconic acid, δ-lactone 35.70 Tripotassium citrate 41.20 Xylitol 48.00 2,2-Dimethyl-1,3-dioxolane-4- methanol 61.50 L-Tartaric acid dipotassium salt 65.60 Potassium acetate 15.30 a Costs based on Sigma-Aldrich. Flash Point The flash point temperature is a measure of the tendency of a material to form a flammable mixture with air. According to the deicing/anti-icing specifications, the flash point of these formulations should not be lower than 100°C as determined by the American Society for Testing and Materials (ASTM) D93 Standard Test Methods for Flash Point by the Pensky- Martens Closed Cup Tester. Table 4-3 presents the flash point temperatures and temperature limits for all of the FPDs with water (1:1 by weight) and as neat solution. With respect to safety considerations, four FPD mixtures having flash points lower than 100°C were rejected for further evaluation. Two of these FPD mixtures were also rejected

ALTERNATIVE AIRCRAFT ANTI-ICING FORMULATIONS 4-8 because they were frozen at -14.5 °C and two were rejected that were frozen at -20 °C. Table 4-5 summarizes the FPD mixtures that met the requirements for freezing point depression and flash point. Eleven FPDs that did not meet the criteria are also listed. TABLE 4-5. Summary of FPD mixture aircraft and runway deicing/anti-icing agents meeting performance and safety requirements. FPD:Water Mixtures Further evaluated for aircraft and runway applications 1,3-Butanediol Diethylene glycol Dipropylene glycol D-Gluconic acid, δ-lactone Glycerol 2-Methyl-1,3-propanediol 1,2-Propylene glycol 1,3-Propylene glycol Sodium acetate L-Tartaric acid dipotassium salt Triethylene glycol Trimethylopropane Tripotassium citrate Further evaluated for runway deicing applications 2,2-Dimethyl-1,3-dioxolane-4-methanol Xylitol Rejected from further evaluation 2,3-Butanediol Calcium propionate Dimethyl malonate Dimethyl succinate Disodium succinate 2-(2-Ethoxyethoxy)-ethanol Ethylene carbonate 4-Methyl-γ-butyrolactone 2-(2-Methyloxyethoxy)-ethanol Propylene carbonate 1,1,1-Trimethanolethane The FPDs listed in Table 4-5 that met the criteria for deicing performance and safety criteria were evaluated for their BOD and COD and aquatic toxicity, as described below.

SECTION 4—3BTIER 1 RESULTS 4-9 Oxygen Demand Methods COD and BOD analyses were conducted on candidate FPDs. COD analyses were performed in triplicate according to ASTM method D1252-88(B). Traditional 5-day BOD analyses were performed in quadruplicate according to Standard Methods for the Examination of Water and Wastewater, method 5210B (31). Twenty-eight-day time-series BOD analyses were performed in triplicate using a modification to method 5210B that used 2,120-mL custom BOD bottles rather than the traditional 300-mL bottles. Single dilutions were prepared based on the results of the COD and conventional BOD5 analyses, and the dissolved oxygen was monitored at 5, 15, and 28 days. As DO concentrations approached 2.0 mg/L, the samples were re-aerated using filtered compressed air and then returned to the incubator. Results were considered valid as long as 2 milligrams or more of oxygen depletion was observed in test vials at each measurement interval during the test. Measurement uncertainty was considered to be too large for accurate reporting with less than 2 mg of oxygen demand exerted over the time interval. COD was used as an estimate of ultimate BOD to compute percent biodegradability of the deicer formulations exerted over time. The liquid and solid deicer products tested in this study were “neat” products and devoid of ammonia or other combined nitrogen compounds that could contribute to nitrification. Any minimal nitrification that may have occurred during the test periods were eliminated through blank correction. Laboratory blanks were tested along with the samples to correct for the contribution of the “seed” material (i.e., source of microorganisms) and any demand exerted by the reagent water used to prepare the samples. Consequently, only total BOD measurements were made, with the assumption that they would be essentially equivalent to carbonaceous BOD (i.e., oxygen demand for carbon source only). Three replicate glucose- glutamic acid controls were also tested in fresh water at 20°C. Results from these controls showed no indication of method bias. Results Testing was conducted for oxygen demand on 24 FPDs (Table 4-6). Results include computed ThOD and COD for all FPDs. One candidate FPD was eliminated before BOD5 testing was conducted and five candidates were eliminated before the BOD time-series testing was conducted. COD results ranged from 341,000 to 1,880,000 milligrams per kilogram (mg/kg) and compared well with theoretical oxygen demand for all candidate FPDs. In comparison, COD of propylene glycol was 1,620,000 mg/kg and potassium acetate (K-Ac) deicer was previously determined to be 629,000 mg/kg expressed as K-Ac (1). BOD5 results could not be obtained for six of the candidate FPDs: 2-methyl-1,3-propanediol, DEG, dimethyl malonate, 1,1,1- trimethanolethane, and triethylene glycol. In four separate attempts, dimethyl malonate, 1,1,1- trimethanolethane, DEG, and triethylene glycol did not degrade during the 5-day test period. Consequently no results could be obtained. BOD28 data for these particular compounds show similarly low BOD in the first 10-20 days. This lag period was also observed with some of the deicer formulations examined in similar testing with commercial deicer products (1).

ALTERNATIVE AIRCRAFT ANTI-ICING FORMULATIONS 4-10 Biodegradability (BOD result expressed as a percent of COD) based on traditional BOD5 results ranged from 0-76 percent depending on the individual FPDs for which BOD5 could be determined. Some of the very low values (<5 percent) are suspected to be unreliable because of an inability of seed organisms to acclimate to the FPD as a food source in a test as short as 5-days. Biodegradability in the 28-day BOD time-series was also widely variable depending on FPD with a range from <1-78 percent at 5-days, <1-87 percent in 15-days, and 11-89 percent in 28- days (Table 4-6). A lag period in BOD exertion for some FPDs is apparent, with low values in 5- and 15-day tests and increased values in 28-day tests. As was the case with the low BOD5 results, this is likely due to an inability of the seed organisms to acclimate in the short time frame. By 28 days, however, results from most tests indicate that the organisms did acclimate and exhibit substantial BOD. It is uncertain whether an initial acclimation of seed organisms for each individual FPD would ultimately change the final 28-day results. Because of the uncertainties of the traditional BOD5 analyses and the 28-day series and the inability to achieve acceptable BOD results for some products, COD was considered a more reliable and consistent parameter for assessing the potential for oxygen demand in the environment. Therefore, COD was ultimately used in the down-selection process to choose FPDs for inclusion in further testing.

SECTION 4—3BTIER 1 RESULTS 4-11 TABLE 4-6. Summary of COD and BOD results for candidate FPDs in order of increasing COD. FPD CAS Number ThODa (g/kg) CODb (g/kg) BOD5c (g/kg) BOD Time-Seriesd (g/kg) Biodegradation as Percent of COD 5 Days 15 Days 28 Days BOD5 5 Days 15 Days 28 Days L-Tartaric acid dipotassium salt 921-53-9 432 341 231 213 239 242 68 62 70 71 Tripotassium citrate 6100-05-6 345 449 309 262 297 301 69 58 66 67 Disodium succinate 150-90-3 592 684 481 466 533 533 70 68 78 78 Sodium acetate 127-09-3 683 747 552 586 653 667 74 78 87 89 2-(2-Methoxyethoxy)-ethanol 111-77-3 1,730 883 —e 24 517 641 — 3 59 73 Ethylene carbonate 96-49-1 908 899 34 5,870 57,600 96,900 4 1 6 11 D-Gluconic acid,δ-lactone 90-80-2 988 976 660 — — — 68 — — — Calcium propionate 4075-81-4 1,120 1,090 823 791 913 936 76 73 84 86 Xylitol 87-99-0 1,160 1,170 585 644 915 979 50 55 78 84 Glycerol 56-81-5 1,220 1,190 810 846 985 1 68 71 83 84 Propylene carbonate 108-32-7 1,250 1,200 33,500 — — — 3 — — — Diethylene glycol 111-46-6 1,510 1,500 NDf 18,500 128 618 — 1 9 41 Triethylene glycol 112-27-6 1,600 1,610 NDf 53,600 398 560 — 3 25 35 1,2-Propylene glycol 57-55-6 1,680 1,620 973 1,020 1,270 1,310 60 63 78 81 1,3-Propylene glycol 504-63-2 1,680 1,640 731 814 1,070 1,190 45 50 65 73 1,1,1- Trimethanolethane 77-85-0 1,730 1,680 <1,200g 6,310 8,060 406 — 0 0 24 2,2-Dimethyl-1,3-dioxolane-4- methanol 100-79-8 1,820 1,780 10,800 — — — 1 — — — Trimethylolpropane 77-99-6 1,910 1,810 <1,200g -17 -21 660 — -1 -1 36 2,3-Butanediol 513-85-9 1,950 1,820 1,200 — — — 66 — — —

ALTERNATIVE AIRCRAFT ANTI-ICING FORMULATIONS 4-12 TABLE 4-6. Summary of COD and BOD results for candidate FPDs in order of increasing COD. FPD CAS Number ThODa (g/kg) CODb (g/kg) BOD5c (g/kg) BOD Time-Seriesd (g/kg) Biodegradation as Percent of COD 5 Days 15 Days 28 Days BOD5 5 Days 15 Days 28 Days 1,3-Butanediol 107-88-0 1,950 1,830 820 843 1,390 1,450 45 46 76 79 2-Methyl-1,3-propanediol 2163-42-0 1,950 1,850 NDe 807 1,450 1,480 — 44 78 80 Dipropylene glycol 25265-71- 8 1,910 1,860 1,580 12 81 935 0 1 4 50 2-(2-Ethoxyethoxy)-ethanol 111-90-0 1,910 1,880 1,100 399 1,490 1,540 59 21 79 82 4-Methyl-γ-butyrolactone 108-29-2 1,920 1,880 814 — — — 43 — — — aTheoretical oxygen demand is computed as ThOD = 32 x number of carbons + 8 x number of hydrogens + 16 x number of oxygens. bThe percent relative standard deviation from 3 replicates was 5% or less for all COD samples except for 3 samples that had only one replicate due to early elimination of FPD consideration (propylene carbonate, 2,3-butanediol, and D-Gluconic acid,δ-lactone). cThe percent relative standard deviation from 4 replicates was 10% or less for all BOD5 samples except 2,2-Dimethyl-1,3-dioxolane-4-methanol, Disodium succinate, Propylene carbonate, and Xylitol. dThe percent relative standard deviation from 3 replicates was 10% or less for all BOD time-series samples except 1,1,1- Trimethanolethane, 2-(2-Methoxyethoxy)-ethanol, 2- Methyl-1,3-propanediol, DEG, Dipropylene glycol, and Trimethylolpropane. f— not determined fBOD5 could not be determined due to suspected inability to find optimum sample to organism ratio. gNo measurable demand from BOD5 analysis.

SECTION 4—TIER 2 RESULTS 4-13 Aquatic Toxicity Methods Screening toxicity tests were used to approximate toxic endpoints. Procedures to screen a large list of potential deicer components were similar to range-finding toxicity tests, which are used to choose the correct range of concentrations prior to a full definitive acute or chronic bioassay. If toxicity data were available from MSDSs, concentrations were prepared based on those data. A 50 percent dilution series of up to 10 dilutions were prepared for the screening assay for each product tested. Five <24-hour-old Ceriodaphnia dubia and two <24- hour-old Pimephales promelas larvae were placed in the same test chamber. For several days following hatching, the fish are too young to consume daphnia. Two replicates were prepared for each product concentration. Test duration was 48 hours without renewal. No water quality parameters were measured during or after screening tests. Microtox® assays were performed without replication. If prepared concentrations were found to be out of range for approximating an LC50 or half maximal effective concentration (EC50) for any species, a new set of concentrations were prepared and the screening test was repeated for those species. LC50s and EC50s reported for screening tests should only be considered approximations because procedures differ from standard definitive bioassays. Screening toxicity procedures include fewer replicates, non-renewal of test solutions, shorter exposure duration for Pimephales promelas, and other procedural variances from definitive toxicity tests. Uncertainty levels cannot be determined due to lack of replication. It is assumed the actual EC or LC50 value lies between the test concentrations which are below and above the calculated EC50 or LC50. For this dilution series, the lower and upper limits are one half and double the EC50 or LC50 respectively. Results Results from screening-level bioassays for 24 FPDs (Table 4-7), 19 surfactants (Table 4-8), 14 corrosion inhibitors (Table 4-9), and 6 thickeners (Table 4-10) are presented below. Previous work has identified surfactants and benzotriazole-based corrosion inhibitors as the primary source of toxicity in previous commercial formulations (1). Therefore, FPDs with similar toxicity characteristics, such as 1,2 propylene glycol, would be acceptable alternatives if oxygen demand were less. Considering only aquatic toxicity, several FPDs could be viable candidates, including DEG, xylitol, glycerol, triethylene glycol, and dipropylene glycol. In addition, 2-(2-methoxyethoxy)-ethanol and 1,3 butanediol could be considered if the oxygen demand characteristics were greatly improved over 1,2 propylene glycol. The evaluation process for candidate surfactants, corrosion inhibitors, and thickeners included consideration of the aquatic toxicity profiles as well as the concentration needed to provide favorable performance. Testing began with aquatic toxicity evaluation, and the results are presented in Tables 4-8, 4-9, and 4-10. These results are only approximations. Compounds are organized by least toxic to most toxic endpoint, determined by the most sensitive species. Evaluation of performance properties for specific concentrations of these formulation components were conducted in Tier 2 and are discussed in Section 5. Multiple candidates for alternative surfactants, corrosion inhibitors, and thickeners indicate a potential for improved toxicity profiles over these components in previous commercial formulations.

ALTERNATIVE AIRCRAFT ANTI-ICING FORMULATIONS 4-14 TABLE 4-7. Screening-level toxicity data for candidate FPDs for three species. The shaded boxes indicate the most sensitive species to each compound. FPD Aircraft or Pavement Microtox® EC50a (mg/L) C. dubia LC50b (mg/L) P. promelas LC50 (mg/L) Diethylene glycol A,P 66,900 53,000 56,900 Xylitol P 348,000 48,500 52,000 1,2-Propylene glycol A,P 69,300 46,300 49,600 Glycerol A,P 130,000 34,900 46,000 Triethylene glycol A,P 54,000 31,000 59,900 Dipropylene glycol A,P 25,200 20,600 35,900 2-(2-Methoxyethoxy)-ethanol A,P 56,700 22,100 18,000 1,3-Butanediol A,P 17,000 35,900 35,900 2-(2-Ethoxyethoxy)-ethanol A,P 13,600 16,600 14,400 1,3-Propylene glycol A,P 29,300 25,300 12,700 1,1,1-Trimethanolethane P 11,200 26,400 35,800 Ethylene carbonate P 24,100 8,850 10,500 2-Methyl-1,3-propanediol A,P 8,530 16,400 32,000 Trimethylolpropane P 7,550 10,000 29,700 Sodium acetate P 44,600 7,350 7,350 sodium formate P 51,800 4,140 7,460 Disodium succinate P 73,300 3,540 11,300 Propylene carbonate A 2,880 12,100 4,420 2,2-Dimethyl-1,3-dioxolane-4- methanol A,P 2,540 8,840 14,900 Calcium propionate P 39,900 2,260 13,800 L-Tartaric acid dipotassium salt P 94,100 859 1,020 potassium formate P 14,700 730 940 Tripotassium citrate P 34,300 365 433 D-Gluconic acid,δ-lactone P 408 212 235 aThe Microtox® EC50 is the statistically determined concentration that would result in a 50% reduction in light emission compared to a laboratory control. bThe LC50 is the statistically determined concentration that would cause death in 50% of the population exposed. Previous work indicated that alkylphenol and alcohol ethoxylate surfactants approximating those in previous deicer formulations had acute toxicity endpoints between 0.44 and 16 mg/L (1). Eleven of the candidate surfactants evaluated in this study show potential for

SECTION 4—3BTIER 1 RESULTS 4-15 improved toxicity profiles (Table 4-8). These surfactants include the first eleven candidates in Table 4-8. Considering the minimum endpoint for the three organisms, values range from 20.7 mg/L (moderately less toxic than current surfactants) to 14,900 mg/L (substantially less toxic than current surfactants). Table 4-8. Screening-level toxicity data for candidate surfactants for three species. The shaded boxes indicate the most sensitive species to each compound. Surfactant Microtox® EC50 (mg/L) C. dubia LC50 (mg/L) P. promelas LC50 (mg/L) Tergitol L-64 25,000 28,200 14,900 Surfynol 465 1,120 686 437 Tetronic 904 402 25,900 7,210 Triton CF-32 715 361 361 Tergitol TMN-10 + 10% Ridafoam 387 183 105 Tergitol TMN-10 408 160 91.9 Lutensol XP 100 118 54.6 56.5 Tergitol TMN-6 140 65.3 39 Triton CG-110 29.8 163 361 Triton CG-110 + 10% Ridafoam 27.6 298 735 Plurafac S-405LF 94.7 40.7 20.7 Lutensol TDA 10 16.1 14.5 25.3 Triton DF-16 15.6 13.1 15.0 Tergitol 15-S-12 153 12.4 12.4 Lutensol XP 50 25.3 9.34 23.2 Tergitol 15-S-7 13.2 5.30 5.30 Bio Soft N1-7 3.60 3.60 10.2 Ridafoam (anti-foaming product) 3.03 22.8 >780 Merpol SE 2.28 9.33 4.67 Bio-Soft N1-5 1.92 2.47 6.60 Acute toxicity endpoints for benzotriazole-based corrosion inhibitors varied between 4.3 and 81 mg/L in previous testing (32). Nine of the candidate corrosion inhibitors evaluated in this study show potential for improved toxicity profiles (Table 4-9). These corrosion inhibitors include the first nine candidates in Table 4-9. Considering the minimum endpoint for the three organisms, values range from 46.3 mg/L (similar toxicity to current corrosion inhibitors, but other organisms are much less sensitive) to 375 mg/L (substantially less toxic than current corrosion inhibitors).

ALTERNATIVE AIRCRAFT ANTI-ICING FORMULATIONS 4-16 TABLE 4-9. Screening-level toxicity data for candidate corrosion inhibitors for three species. The shaded boxes indicate the most sensitive species to each compound. Corrosion inhibitor Microtox® EC50 (mg/L) C. dubia LC50 (mg/L) P. promelas LC50 (mg/L) Mazon RI 325 375 481 8,840 Triethanolamine 212 1,430 11,500 Sodium silicate 273 223 208 Potassium silicate 164 445 445 Sodium borate decahydrate 150 836 1,780 Potassium phosphate 102 253 311 Korantin SMK 48.0 61.6 119 Potassium carbonate 46.3 310 191 3-Methoxypropylamine 33.5 76.5 101 Cobratec 948 10.4 354 177 Ammonyx CDO Special 35.2 39.2 8.03 DrewPlast 154 3.75 11.1 8.84 Ninol 1301 86.4 3.54 3.54 Ninol 201 38.0 1.06 3.01 Aquatic toxicity endpoints of polyacrilic acid thickeners (Carbopol®) found in commercial products were less (more toxic) than other candidate thickeners. However, given the concentrations used in Type IV formulations, results did not warrant dismissal of any candidate thickeners based solely on toxicity (Table 4-10). Candidates, therefore, must be selected for further testing based upon performance and cost. TABLE 4-10. Screening-Level Toxicity Data for Candidate Thickeners for Three Species Thickener Microtox® EC50 (mg/L) C. dubia LC50 (mg/L) P. promelas LC50 (mg/L) Cellosize DCS HV 7,290 2,770 6,360 Kelzan-HP 2,440 1,310 2,380 Kelzan-RD 2,480 862 2,120 K1A96 2,230 853 3,180 Carbopol EZ-4 (neutralized with KOH) 793 154 177 Carbopol EZ-4 (neutralized with TEA) 375 50.0 265

SECTION 4—3BTIER 1 RESULTS 4-17 In addition, tripotassium citrate and potassium carbonate were evaluated as anti-caking agents for sodium formate granular pavement deicer material. Aquatic toxicity was evaluated for tripotassium citrate as a candidate FPD (Table 4-7). Aquatic toxicity endpoints for potassium carbonate were as follows: Microtox® EC50, 48.4; Ceriodaphnia dubia LC50, 277 mg/L; Pimephales promelas LC50, 277 mg/L. Considering the low concentrations needed as anti-caking agents, neither compound was dismissed because of aquatic toxicity results. Viscosity Viscosity measurements of six candidate thickeners for Type IV aircraft anti-icing formulations were made using a Brookfield viscometer. The thickeners were mixed with water at different concentrations, and the viscosity was determined at room temperature (20°C) and at 5°C and compared to a commercially available Type IV anti-icing formulation. The concentration for each thickener in water was based on values obtained in the literature. The test results are presented in Figure 4-2 for each surfactant at the two temperatures. The viscosity curves at both temperatures are virtually parallel to each other; the values are within 20 percent of each other at the two temperatures. Based on this observation, subsequent viscosity measurements were made at room temperature. A comparison of the viscosities of the surfactant:water mixtures with a commercially available Type IV anti-icing formulation is shown in Figure 4-3. Table 4-11 summarizes the performance testing results of the thickeners and shows the change in thickener concentration to match the viscosity of the commercial Type IV formulation (see each series of curves in Figure 4-3). The viscosity of the Cellulose DCS HV thickener was 37 times lower than the commercial formulation and is not shown in Figures 4-2 and 4-3. TABLE 4-11. Performance testing results for thickeners. Thickener Test Concentration, wt % Concentration Changea Kelzan HP 0.4 Same Kelzan RD 0.5 Increase K1A96 0.75 Decrease Cellulose DCS HV 1.5 Substantial increase Carbopol EZ-4 with TEA 0.1 Decrease Carbopol EZ-4 with KOH 0.1 Decrease aChange in concentration indicates the direction of the test concentration to match the viscosity of the commercially available Type IV runway anti-icer, e.g., the concentration of Kelzan RD must be increased from 0.5 wt% to match the concentration of the commercial Type IV anti-icer.

ALTERNATIVE AIRCRAFT ANTI-ICING FORMULATIONS 4-18 Figure 4-2. Thickener viscosity/shear rate at 5°C and at room temperature. Kelzan HP (0.5%) 10 100 1000 10000 100000 1000000 0.01 0.1 1 10 100 Shear Rate, 1/sec Vi sc os ity , c P 20ºC 5ºC Kelzan RD (0.05%) 0.01 0.1 1 10 100 Shear Rate, 1/sec 20ºC 5ºC K1A96 (0.75%) 100 1000 10000 100000 1000000 0.01 0.1 1 10 100 Shear Rate, 1/sesc Vi sc os ity , c P 20ºC 5ºC Carbopol EZ-4 (0.1%) 0.01 0.1 1 10 100 Shear Rate, 1/sec 20ºC TEA pH=8 5ºC TEA pH=8 20ºC KOH pH=8 Commercial Type IV 100 1000 10000 100000 1000000 0.01 0.1 1 10 100 Shear Rate, 1/sec Vi sc os ity , c P 20ºC 5ºC

SECTION 4—3BTIER 1 RESULTS 4-19 Figure 4-3. Comparison of viscosity for surfactants with commercially available Type IV anti-icing formulation at room temperature. Comparison of Kelzan HP and Commercial Type IV 10 100 1000 10000 100000 0.01 0.1 1 10 100 Shear Rate, 1/sec Vi sc os ity , c p Kelzan HP (0.5 %) Commercial Type IV Comparison of Kelzan RD and Commercial Type IV 0.01 0.1 1 10 100 Shear Rate, 1/sec Kelzan RD (0.5 %) Commercial Type IV Comparison of K1A96 and Commercial Type IV 10 100 1000 10000 100000 1000000 0.01 0.1 1 10 100 Shear Rate, 1/sec Vi sc os ity , c p Commercial Type IV K1A96 (0.75%) Comparison of Carbopol EZ-4 and Commercial Type IV 0.01 0.1 1 10 100 Shear Rate, 1/sec Commercial Type IV Carbopol EZ-4 (pH=8) Contact Angle The contact angle is a measure of the degree to which the droplet will spread over the surface—the lower the contact angle between the water:surfactant droplet and the surface, the lower the surface tension, and the more the droplet will spread over a wing surface. The contact angle of two concentrations of surfactant:water mixtures (0.5:99.5 and 1:99 percent by weight) was determined using a drop shape analyzer. The measurements were made at the University of Massachusetts-Lowell on a Krüss Drop Shape Analysis System DSA100. Test results are shown in Table 4-12 for the surfactant:water mixtures. The contact angle, a measure of surfactant effectiveness, is generally the same at the two concentrations.

ALTERNATIVE AIRCRAFT ANTI-ICING FORMULATIONS 4-20 TABLE 4-12. Performance testing results for surfactants. Surfactant Contact angle, degrees Ranking by contact angle 0.5:99.5 1:99 Average Tergitol L-64 31.5 28.7 30.1 8 Tetronic 904 26.2 23.8 25.0 7 Mixture 2a 13.0 13.0 13.0 4 Surfynol 465 13.1 9.5 11.3 3 Triton CF-32 18.4 22.3 20.4 6 Triton CG-110 19.7 14.3 27.0 5 Mixture 1b 8.9 11.3 10.1 2 Tergitol TMN-10 7.1 8.5 7.8 1 Distilled water 64.2 64.2 aMixture 2: Triton CG-110+10% Ridafoam NS 221. bMixture 1: Tergitol TMN10+10% Ridafoam NS 221. Down-Selection of Components for Tier 2 Testing FPDs Down-selection of FPDs included consideration of results from tests for aquatic toxicity, COD, freezing point, flash point, and a qualitative observation of miscibility (Table 4-13). Aquatic toxicity criteria of 20,000 mg/L for aircraft deicing and anti-icing fluids and 2,000 mg/L for pavement deicer material were used. Because of the uncertainties involved in BOD testing, it was decided that COD would be used in the down-selection process as a more reliable and consistent measure of oxygen demand. COD criteria of 1,600,000 mg/kg (the value for 1,2 propylene glycol) for aircraft deicing and anti-icing fluids and 629,000 mg/kg (the value for potassium acetate) for pavement deicer material were used. The freezing point for FPDs needed to be less than -14.5°C for aircraft deicing and anti-icing fluids and less than -20°C for pavement deicing material to be considered further. FPDs needed a flashpoint of greater than 100°C to be considered further. All FPDs needed to be completely miscible in water in order to be considered further. These eliminating criteria left only two FPDs for aircraft deicing and anti-icing fluids: glycerol and diethylene glycol. No fluids met test criteria to continue testing for new pavement deicing materials.

SECTION 4—3BTIER 1 RESULTS 4-21 TABLE 4-13. Down-selection results for candidate FPDs for potential use as aircraft and pavement deicers and anti-icers. Freezing Point Depressants Elimination Factors for Selection of Candidate Deicers and Anti-Icers Aircraft Pavement Liquid 2,3-Butanediol Freezing Pt., COD 4-Methyl-γ-butyrolactone Freezing Pt., Toxicity Dimethyl malonate Immiscible, Freezing Pt. Aircraft only Dimethyl succinate Immiscible, Freezing Pt. Propylene carbonate Freezing Pt, Toxicity 1,2-Propylene glycol Current-use Current-use 1,3-Butanediol Toxicity, COD COD 1,3-Propylene glycol Toxicity, COD COD 2-(2-Ethoxyethoxy)-ethanol Toxicity, COD COD 2,2-Dimethyl-1,3-dioxolane-4- methanol Flash pt., Freezing Pt., Toxicity Flash pt. 2-Methyl-1,3-propanediol Toxicity, COD COD Dipropylene glycol COD COD Glycerol Continue testing COD Triethylene glycol COD COD 2-(2-Methoxyethoxy)-ethanol Flash pt., toxicity Flash pt. Diethylene glycol Continue testing COD Solid 1,1,1- Trimethanolethane COD Calcium propionate Freezing Pt. COD D-Gluconic acid,δ-lactone Toxicity, COD Disodium succinate Freezing Pt., COD Ethylene carbonate Freezing Pt., COD L-Tartaric acid dipotassium salt Pavement only Toxicity Potassium L-lactate Not available Sodium acetate Current-use Sodium formate Current-use Trimethylolpropane COD Tripotassium citrate Toxicity Xylitol COD

ALTERNATIVE AIRCRAFT ANTI-ICING FORMULATIONS 4-22 Thickeners The thickeners were down-selected based on the aquatic toxicity and changes in the test concentration of the thickener necessary to match the viscosity of the commercially available Type IV aircraft deicing formulation. Table 4-14 shows the outcome of the down-selection of the thickeners. TABLE 4-14. Down-selection rankings for candidate thickeners. Thickener Test Concentration, wt% Pimephales promelas LC50 (mg/L) Change in Pimephales promelas LC50a Rank K1A96 0.75 424,000 Increase 1 Kalzan HP 0.5 476,000 Increase 2 Carbopol EZ-4 with TEA 0.1 106,000 Increase 3 Kalzan RD 0.5 424,000 Decrease 4 Carbopol EZ-4 with KOH 0.1 71,000 Increase 5 Cellulosize DCS HV 1.5 424,000 Substantial decrease 6 a Resulting from changing thickener concentration from the test concentration to match viscosity of commercially available Type IV aircraft anti-icer. Change in Pimephales promelas LC50 is in the direction opposite to the concentration change in Table 4-11. For example, increasing the concentration of Kalzan HP from 0.5 wt% to match the concentration of the commercial Type IV anti-icer will reduce the LC50 and increase the toxicity of the solution. Surfactants The surfactants were ranked on the basis of their aquatic toxicity and effectiveness based on contact angle. Results indicate that toxicity of these surfactants generally increases as surfactant performance improves (Table 4-15). TABLE 4-15. Performance and aquatic toxicity testing results for surfactants. Surfactant Pimephales promelas LC50 (mg/L) Contact Angle, degrees Ranking by Contact Angle 0.5:99.5 1:99 Average Tergitol L-64 14895 31.5 28.7 30.1 8 Tetronic 904 7212 26.2 23.8 25.0 7 Mixture 2a 725 13.0 13.0 13.0 4 Surfynol 465 437 13.1 9.5 11.3 3 Triton CF-32 361 18.4 22.3 20.4 6 Triton CG-110 361 19.7 14.3 27.0 5 Mixture 1b 105 8.9 11.3 10.1 2 Tergitol TMN-10 92 7.1 8.5 7.8 1 Distilled water 64.2 64.2 aMixture 2: Triton CG-110+10% Ridafoam NS 221. bMixture 1: Tergitol TMN10+10% Ridafoam NS 221.

SECTION 4—3BTIER 1 RESULTS 4-23 Corrosion Inhibitors Two corrosion inhibitors, TEA and Mazon RI 235, were down-selected for further evaluation in Tier 2, based on aquatic toxicity (Table 4-13).