Improving sustainable industrial wastewater treatment with ion exchange of NTO

As the complexity of industrial wastewater increases, there is an increasing need for research in water treatment technologies in order to properly treat industrial wastewater in cost effective and sustainable ways. The U.S. military is currently utilizing insensitive high explosives (IHEs) containing NTO (3-nitro-1,2,4-triazol-5-one) to supplement conventional and less safe munitions, such as TNT. The work presented herein investigates the treatment of NTO contaminated wastewater with anion exchange resins. More specifically, this work investigates: (1) the use of nitrate and perchlorate selective anion exchange resins for exchanging NTO for chloride (2) the use of brine solution to regenerate such resins (3) the kinetics of NTO exchange by such resins and (4) the use of anion exchange resins for exchanging NTO when nitrate is in solution. Using high performance liquid chromatography with ultraviolet light detection (HPLC - UV), NTO was quantified before and after batch studies, showing that all four anion exchange resins investigated (two nitrate and two perchlorate) were capable of exchanging NTO for chloride and were able to be regenerated with brine solution. Time constraints limited the other studies to test the performance associated with only one of the four resins, Amberlite PWA5, a macroporous, nitrate selective, anion exchange resin from DOW. Regarding kinetics, Amberlite PWA5 exchanged NTO for chloride over a 24 hour period. Additionally, with nitrate in solution, Amberlite PWA5 still exchanged NTO for chloride. Ultimately, results demonstrate that ion exchange technology appears to be a suitable alternative treatment technology for NTO-contaminated wastewater.

IMPROVING SUSTAINABLE INDUSTRIAL WASTEWATER TREATMENT WITH ION EXCHANGE OF NTO by Dana Tran A Senior Honors Thesis Submitted to the Faculty of The University of Utah In Partial Fulfillment of the Requirements for the Honors Degree in Bachelor of Science In Civil and Environmental Engineering Approved: ______________________________ Jennifer L. Weidhaas PhD, PE Thesis Faculty Supervisor _____________________________ Michael E. Barber, PhD, PE Chair, Department of Civil and Environmental Engineering _______________________________ Michael E. Barber, PhD, PE Honors Faculty Advisor _____________________________ Sylvia D. Torti, PhD Dean, Honors College April 2019 Copyright © 2019 All Rights Reserved ABSTRACT As the complexity of industrial wastewater increases, there is an increasing need for research in water treatment technologies in order to properly treat industrial wastewater in cost effective and sustainable ways. The U.S. military is currently utilizing insensitive high explosives (IHEs) containing NTO (3-nitro-1,2,4–triazol–5-one) to supplement conventional and less safe munitions, such as TNT. The work presented herein investigates the treatment of NTO contaminated wastewater with anion exchange resins. More specifically, this work investigates: (1) the use of nitrate and perchlorate selective anion exchange resins for exchanging NTO for chloride (2) the use of brine solution to regenerate such resins (3) the kinetics of NTO exchange by such resins and (4) the use of anion exchange resins for exchanging NTO when nitrate is in solution. Using high performance liquid chromatography with ultraviolet light detection (HPLC - UV), NTO was quantified before and after batch studies, showing that all four anion exchange resins investigated (two nitrate and two perchlorate) were capable of exchanging NTO for chloride and were able to be regenerated with brine solution. Time constraints limited the other studies to test the performance associated with only one of the four resins, Amberlite PWA5, a macroporous, nitrate selective, anion exchange resin from DOW. Regarding kinetics, Amberlite PWA5 exchanged NTO for chloride over a 24 hour period. Additionally, with nitrate in solution, Amberlite PWA5 still exchanged NTO for chloride. Ultimately, results demonstrate that ion exchange technology appears to be a suitable alternative treatment technology for NTO-contaminated wastewater. i TABLE OF CONTENTS ABSTRACT......................................................................................................................... i 1 INTRODUCTION ...................................................................................................... 1 2 MATERIALS AND METHODS ................................................................................ 6 2.1 Materials .............................................................................................................. 6 2.2 Experimental Procedures ..................................................................................... 7 2.2.1 Ion Exchange Method ................................................................................... 7 2.2.2 Regeneration Method .................................................................................... 8 2.2.3 Complex Solution Ion Exchange Method ..................................................... 8 2.2.4 Kinetic Ion Exchange Method ...................................................................... 9 2.3 3 4 Analytical Methods ............................................................................................... 9 2.3.1 NTO Concentration Analysis ........................................................................ 9 2.3.2 HPLC – NTO Standard Curves................................................................... 10 2.3.3 Spectrophotometer – Nitrate Standard Curve ............................................. 11 RESULTS ................................................................................................................. 12 3.1 Kinetic Study Results .......................................................................................... 12 3.2 Ion Exchange and Regeneration Study Results .................................................. 14 3.3 Complex Solution Ion Exchange and Regeneration Study Results .................... 16 DISCUSSION ........................................................................................................... 18 4.1 Variability of Exchange Capacity Results .......................................................... 18 4.2 Batch versus Column Study Considerations ...................................................... 20 4.3 Industrial Scale IX System Consideration.......................................................... 21 5 CONCLUSION ......................................................................................................... 22 6 APPENDICES .......................................................................................................... 23 6.1 Appendix I: Additional Standard Curve Equations............................................ 23 6.2 Appendix II: Exchange Capacities ..................................................................... 23 6.2.1 Appendix II - A: Theoretical Exchange Capacities .................................... 23 6.2.2 Appendix II - B: Ion Exchange and Regeneration Exchange Capacities ... 24 6.2.3 Appendix II - C: Complex Solution Ion Exchange and Regeneration Exchange Capacities ................................................................................................. 25 7 ACKNOWLEDGEMENTS ...................................................................................... 27 8 REFERENCES ......................................................................................................... 28 1 1 INTRODUCTION The U.S. military is currently utilizing insensitive high explosives (IHEs) containing DNAN (2,4-dinitroanisole, NTO (3-nitro-1,2,4–triazol–5-one), and NQ (nitroguanidine) to supplement conventional and less safe munitions, such as TNT. In particular, NTO (Figure 1) performs similarly to the well-known explosive, RDX (1,3,5-trinitro-1,3,5triazine), but has greater resilience to physical insults (such as heat, impact, and friction) and does not detonate during accidental ignition [1]. These properties fulfill the required safety parameters of an IHE, which encourage the development of NTO-based formulations to an industrial-scale [1]. The implications associated with such growth is that in the process of creating IHEs, munition manufacturers, load assemble pack (LAP) facilities, and demilitarization facilities generate wastewater streams that contain concentrations of IHEs. Figure 1 – Structure of NTO Wastewater containing NTO is a challenge for existing treatment plants for several reasons. First, NTO is a highly soluble compound (1.26g/100mL at 19° C) [2], thus the wastewater produced during the manufacture and processing of NTO can be heavily concentrated with NTO. Up to 10-15 kg of dissolved NTO may be present in 2 1000 L of manufacturing plant wastewater from the melt-pour process. [3]. Thus, aggressive or longer treatment times more likely necessary. Second, aqueous NTO is acidic (approximate pKa = 3) [4] , which creates concerns about the corrosivity of the NTO contaminated wastewater to treatment systems. Third, NTO is less likely to respond to traditional treatment methods, such as sorption to granular activated carbon (GAC) [4], as its negative charge in aqueous solution make it less likely to sorb to carbon. In regards to research into alternative treatment methods, the majority of current research is focused on chemical treatment of IHEs, but these processes are typically not cost-effective and can produce a number of unknown daughter compounds in the process[5]. For example, a study investigating bioremediation for the treatment of DNAN (a type of IHE), had results that were generally not very successful [5]. DNAN was subjected to biological transformation in an anaerobic bioreactor, but the microorganisms produced secondary compounds that were as toxic and harmful as the parent compound [6], highlighting the need for further research into alternative treatment solutions for IHEs. Ion exchange technology is an alternative solution for treating NTO compared traditional treatment methods, such as GAC. As mentioned previously, the ionic nature of NTO in solution makes GAC less likely to be an effective treatment method, whereas ion exchange utilizes it. In addition, ion exchange resins may have a longer lifespan compared to GAC, as resins can be re-used through a regeneration process. However, the implication associated with regeneration involves the recharge solution (brine), which will contain NTO after the chlorides in solution exchange for NTO on the resin. One study investigated the abiotic degradation of NTO via zero valent iron and Fenton’s 3 Reagent to some success, which may be a viable treatment solution that follows regeneration, but the solution has yet to be investigated at full scale [7]. Herein, it is proposed that ion exchange (IX) may be a viable treatment for wastewater containing NTO. The exchange process is conducted by ion exchange resins, which are small, polymer microbeads that contain permanently attached fixed ions. To preserve the electrical neutrality of the resin, the microbead also contains a mobile counterion that can move in and out of the bead. For example, an anion resin bead structure may have polystyrene (C8H8) linked with divinylbenzene (C10H10) as the permanently attached fixed ions, with chlorides as the mobile ions (Figure 2). In the case of anion exchange resins, when contaminated solution meets a resin bead, a contaminant anion that goes into the bead will be replaced by the chloride leaving the bead to enter the solution. This process is ion exchange. Figure 2 – Schematic anion exchange resin bead. Modified figure from LennTech [8]. 4 Like other treatment materials, the resins become exhausted over time. Ion exchange resins can be re-used by the process of regeneration, which reverses the exchange process with brine solution. The chlorides in a heavily concentrated brine solution will be exchanged into the resin, taking place of the original contaminant ion and releasing it back into solution. On a large scale, a contaminated solution would be treated by being exposed to a bed of IX resins, where industrial columns facilitate the exchange and regeneration processes (Figure 3). Figure 3 – Ion Exchange Industrial Column. Modified figure from Dardel [9]. Ion exchange takes advantage of the properties of NTO. As an anion in neutral pH water, NTO will exchange with chlorides from the anion exchange resin. Taking into consideration wastewater conditions, the contaminant solution will likely contain other cations and anions, which will affect the NTO exchange performance of the resin [10]. The performance of an anion exchange resin will likely decrease when treating NTO 5 contaminated solution with other anion loads as there is more competition between anions for exchange sites on the resin [10]. NTO does not have a large affinity compared to the other potential ions in solution, such as nitrate or sulfate. On the other hand, the performance of the resin when treating NTO contaminated solution with other cation loads should remain relatively similar or unaffected, as cations do not have an affinity for the resin charge but may interfere with the pathway between NTO and the resin exchange sites [10] . Munitions manufacturing wastewater will contain other ions and munitions (such as nitrate, perchlorate, DNAN and NQ) [11] [12], thus the exchange capacity of the resin will depend on the concentrations of the types of ion present in solution. The majority of research on ion exchange has been related to removal of hazardous organics from drinking water, but there is limited research exploring the removal of explosives in industrial wastewater treatment, especially the removal of IHEs from wastewater streams. As such, this work involves studies that investigates: (1) the use of nitrate and perchlorate selective anion exchange resins for exchanging NTO for chloride (2) the use of brine solution to regenerate such resins (3) the kinetics of NTO exchange by such resins and (4) the use of anion exchange resins for exchanging NTO when nitrate is in solution. The exchange of NTO by anion exchange resins was investigated in batch experiments using four different anion exchange resins (Table 1). Nitrate and perchlorate selective anion exchange resins were chosen because of the high concentrations of nitrate and perchlorate found alongside NTO in existing munition wastewater streams [11][12]. Nitrate and perchlorate are not only common water contaminants, but they are also byproducts of the munitions manufacturing process[11], thus investigating treatment of 6 nitrate and perchlorate could effectively treat NTO in the process. Additionally, existing vendors of inexpensive anion exchange resins that are nitrate and perchlorate selective were already available made investigating nitrate and perchlorate selective anion exchange resins simpler. Table 1: Anion Exchange Resins Used In Evaluating NTO Removal From Wastewater Resin Vendor Selectivity Resin Matrix Exchange Capacity (eq/L) Moisture Capacity (%) Amberlite PWA5a DOW Nitrate Cross linked copolymer, macroporous >1 52-58 A530Eb Purolite Perchlorate Cross linked copolymer, macroporous 0.6 49-55 PGW6002Ec Purolite Nitrate Cross linked gel copolymer 1.65 40-45 A532Ed Purolite Perchlorate Cross linked gel copolymer 0.85 36-45 References (a) [14], (b) [15], (c) [16], (d) [17] 2 MATERIALS AND METHODS 2.1 Materials The NTO used in these studies was provided by Picatinny Arsenal and was greater than 95% purity. Aqueous NTO stock solutions with concentrations of 1 mg/L to 1000 mg/L were prepared for all studies. The concentrations of NTO stock solutions were validated by NTO standard curves generated from HPLC data before experimentation (Table 2). 7 Aqueous brine (NaCl) stock solutions with concentrations of 12% to 30% were prepared for regeneration studies. The concentrations of NaCl stock solution were not validated during experimentation. Future studies will evaluate the electrical conductivity (EC) of brine solutions to verify brine concentrations. Resins were prepared by saturating 0.05g of dry resin with 1 mL of DI water for 24 hours on a shaker, with 0.9mL of the supernatant decanted before experimentation. Resins used in these studies are listed in Table 1. Preliminary ion exchange and regeneration studies were conducted in 50 mL Falcon plastic centrifuge tubes. Complex solution and kinetic ion exchange studies were conducted in 40 mL amber glass bottles. The transition from using plastic centrifuge tubes to amber glass bottles for later NTO studies occurred because the amber glass bottles did not exhibit significant NTO loss (compared to the plastic centrifuge tube), and was simpler to use to prevent the photodegradation of NTO during experimentation. All samples were filtered (Nylon, 0.45 uM Watman Filters, Fisher Scientific) before HPLC processing. Acetonitrile (Fisher, HPLC gradient grade) and DI water (Fisher, HPLC gradient grade) were the solvents used in HPLC sample processing for all studies. Details about HPLC processing are included in the Analytical Methods section. 2.2 Experimental Procedures 2.2.1 Ion Exchange Method Concentrations of NTO in DI water varied from 1 to 500 mg/L. Controls included solutions with no resin to evaluate losses of NTO by Falcon tubes or Amber bottles. In this study, all four resins in Table 1 were used. The resins were exposed to 30 mL of 8 NTO solution, and samples were taken at time zero and after 24 hours. In the time between sampling, samples were shaken at 120 RPM in a dark space at room temperature. For sampling, 2 mL of solution was syringe filtered, and then placed into amber liquid chromatography (LC) vials. When concentrations of NTO were greater than 100 mg/L, samples were diluted 1:10 with DI water prior to analysis in HPLC. All sample vials were stored at 4 degrees Celsius until analysis by HPLC. Samples from the ion exchange study indicated how much NTO was exchanged for chloride on the resins during the 24 hour period. 2.2.2 Regeneration Method Immediately after the four resins had undergone an exchange period of 24 hours with NTO solution, the supernatant was decanted and the resins were retained. The resins were then exposed to 30 mL of 30% brine solution for 24 hours. Samples were taken at time zero and after 24 hours. In the time between samplings, samples were shaken at 120 RPM in a dark space at room temperatures. At sampling times, 2 mL of solution was syringe filtered, and then placed into amber liquid chromatography (LC) vials. When the resins were exposed to concentrations of NTO were greater than 100 mg/L during the ion exchange study, samples were diluted 1:10 with DI water prior to analysis in the HPLC. All sample vials were stored at 4 degrees Celsius until analysis by HPLC. Samples from the regeneration study indicated how much NTO was released back into the supernatant during the 24 hour period as the resin exchanged NTO for chloride. 2.2.3 Complex Solution Ion Exchange Method To evaluate the potential competition between nitrate and NTO, nitrate was added to NTO solutions for the ion exchange and regeneration studies. Stock solutions had NTO 9 at 300 mg/L and nitrate concentrations varying from 1 to 1000 mg/L. Procedures for the ion exchange and regeneration studies were followed for the complex solution study, apart from the difference in NTO and nitrate concentrations. Experiments were conducted in exclusively amber bottles. Samples from the complex solution study indicated the resin’s exchange capacity for NTO when competing ions are present in solution. Only one resin was tested to examine competition between nitrate and NTO, which was Amberlite PWA5, a macroporous, nitrative selective resin manufactured by DOW. 2.2.4 Kinetic Ion Exchange Method For the kinetic ion exchange study, the ion exchange study was replicated with a higher frequency of sampling events during a 24 hour period. The concentration of NTO tested was assumed to be 500 mg/L. The study examined only one resin, Amberlite PWA5. Sample times during the 24 hour period were: 0, 5, 10, 15, 30, 60, and 120 minutes. Samples were taken from one amber bottle containing the batch study over time (implications associated with this method are discussed in the Results section). 2.3 Analytical Methods 2.3.1 NTO Concentration Analysis The concentration of NTO in solution was measured by LC on a Shimadzu LC – 2030 (Columbia, MD) equipped with a UV detector. Stationary phase was a Synergi TM Hydro-RP (250 x 4.6 mm ID with 4 μm particles, Phenomenex, Torrance, CA). The mobile phase consisted of 95% acetonitrile and 5% water at a flow rate of 0.75 mL/min with 10 uL sample injection volumes for exchange samples. The mobile phase gradient during the 6 minutes was isocratic flow of 95% acetonitrile and 5% water. The retention 10 time of NTO was approximately 2.8 minutes. The monitored wavelength for the method was 348 nm, with a column compartment temperature maintained at 40 degrees Celsius. For regeneration samples, a different mobile phase was used due to the high salt content in order to prevent precipitation of salt crystals in the stationary phase column (i.e. salting out in non-aqueous phase acetonitrile). The regeneration mobile phase consisted of 95% water and 5% acetonitrile. The mobile phase during the 13 minute run consisted of 95:5 % water:acetonitrile for 4 minutes, 1 minute ramp to 5:95% water:acetonitrile, 4 minutes at 5:95% water:acetonitrile, 1 minute ramp to 95:5 % water:acetonitrile and 3 minutes stabilization at 95:5 % water:acetonitrile. The monitored wavelength for the method was 348 nm, with a column compartment temperature maintained at 40 degrees Celsius. The observed retention time of NTO was 3.24 minutes with the HPLC method used to run samples with high salt content, instead of 2.8 minutes with the HPLC method run for samples without salt. 2.3.2 HPLC – NTO Standard Curves For samples that contained NTO exclusively, standard curves were generated by creating solutions of varying NTO concentrations ranging from 0 to 40 mg/L in DI water. Results were plotted and fit to linear regression, shown as Equation 1 in Table 2. Equation 1 was used for all samples that only contained NTO. For samples that contained NTO and NaCl, standard curves were generated by creating solutions of varying NTO and NaCl concentrations in DI water. NTO ranged from 0 to 500 mg/L and NaCl ranging from 0% to 30%. Solutions were diluted 1:10 11 where necessary to limit the amount of NTO injected onto the stationary phase column. Results were plotted and fit to linear regression. Resulting curve equations for 25% and 30% NaCl solutions are shown as Equation 2 and 3 in Table 2, as these were the standard curves were used for all samples that contained NTO and NaCl. The detection range for Equation 2 and 3 is 0.5 – 50 mg/L NTO. Standard curve equations associated with other salt concentrations can be found in Table 7 in Appendix I. Table 2: HPLC Standard Curve Equations for NTO and NaCl Solutions Standard Curve Equation R2 value NTO 1. CNTO = 6E-05(AreaLC) + 0.1923 0.9996 NTO – 25 % NaCl 2. CNTO = 0.0005(AreaLC) – 1.3888 1 NTO – 30% NaCl 3. CNTO = 0.0002(AreaLC) + 0.2403 0.995 It was assumed that the results that contained NTO and nitrate (from the complex solution study) could be interpreted with Equation 1. Resulting chromatograms associated with NTO and nitrate solutions from the HPLC did not show any new peaks or disruptions in the peak associated with NTO, thus it was assumed nitrate was not retained on the HPLC column (C18). 2.3.3 Spectrophotometer – Nitrate Standard Curve Nitrate concentrations were measured by a Hach spectrophotometer using Method 8039. A nitrate standard curve was generated by creating nitrate solutions with concentrations ranging from 0 to 1000 mg/L NaNO3-. Nitrate solutions were diluted 1:10, and tested by spectrophotometer when necessary. Results were plotted and fit to linear regression. Resulting standard curve is shown as Equation 4 in Table 3. 12 Another standard curve was generated for nitrate and NTO solutions for the spectrophotometer. Solutions with 300 mg/L NTO and varying nitrate concentrations from 0 to 1000 mg/L NaNO3- were used. Solutions were diluted 1:10, and tested by spectrophotometer. Results were plotted and fit to linear regression. Resulting standard curve is shown as Equation 5 in Table 3. The detection range for Method 8039 is 0.3 mg/L to 30 mg/L NO3--N. Table 3: Spectrophotometer Standard Curve Equations for NTO and NO3Solutions Standard Curve Equation R2 value NO3- 4. CNO3- = 92.895(AreaSPEC) – 69.25 0.9999 NTO and NO3- 5. CNO3- = 85.997(AreaSPEC) – 56.429 0.995 3 RESULTS 3.1 Kinetic Study Results Amberlite PWA5 resin exchanges NTO with chloride within the first hour of being exposed to NTO solution. Over the 24 hour period, the exchange of NTO reaches a limit and subsequently decreases. The results of the kinetic study with samples taken up to 60 minutes demonstrate that the PWA5 exchange of NTO gradually increases with time, as described by the function in Figure 1. The results of Figure 1 are included in Figure 2, combing a sample point taken at 24 hours to demonstrate kinetic trends over 24 hours. This 24 hour sample point is taken from the ion exchange study results, and is separate from the data collected during the kinetic study itself. 13 NTO Exchanged (mg/L) NTO and Amberlite PWA5 Resin Kinetics 180 160 140 120 100 80 60 40 20 0 y = 0.0461x2 - 0.7645x + 43.521 R² = 0.9557 0 10 20 30 40 50 60 70 Time (minutes) Exchange vs. Time Poly. (Exchange vs. Time) Figure 4 – NTO Kinetic Study Results, (1 HR). Polynomial trendline with equation shown. Figure 5 – NTO Kinetic Study Results (24 HRS). Includes T= 24 sample from separate experiment (T=24 sample point outlined on figure). Exchange of NTO over the 24 hour period reaches max exchange capacity and decreases. The exchange results over 24 hours demonstrates that the resin increases its exchange capacity, reaches a max exchange capacity, and decreases its exchange capacity for NTO. Unexpectedly, the amount of NTO exchanged at zero minutes was greater than 14 the amount of NTO exchanged at five minutes. This is likely not representative of the actual exchange process at time zero. The result may be attributed to a lack of thorough mixing of resin and NTO solution during the time zero sample. Additionally, for each sample event during the kinetic study, 2 mL of solution was removed from the original batch container. With each sampling event, the amount of solution in the batch container decreased, ultimately removing 14 mL of the original 30 mL of NTO solution exposed to the resins. By the end of the study, approximately half of the NTO that was initially in the batch container was removed, effectively reducing the amount of available NTO that could be exchanged by the resins over the 24 hour period. As such, the results may not be representative of the exchange kinetics of the resin, thus future studies will increase the volume of NTO solution to be exposed to the resins. It should also be noted that key samples were not taken in the kinetic study: (1) initial concentrations of NTO solutions were not measured prior to exposure to the resins (instead, they were assumed to be 500 mg/L NTO) and (2) control samples to examine possible changes in NTO concentration influenced by the amber glass bottle. These samples are important to the identifying the kinetics of NTO exchange by the resin. Additionally, replicate studies of the kinetics of the resin were not conducted. Thus, future kinetic studies will be replicated and include samples for initial concentration of NTO solution and control. 3.2 Ion Exchange and Regeneration Study Results The ion exchange and regeneration experiments were intended to examine whether or not ion exchange resins were capable of removing NTO from solution. The ion exchange study indicated that all resins tested were capable of exchanging NTO for 15 chloride during a 24 hour period. The regeneration study indicated that all resins tested were capable of releasing NTO back into solution in exchange for chloride. The exchange (meq/g) associated with each process and for each resin are shown in Table 4, with supporting calculations in Appendix II – B. Table 4: NTO Exchange by Anion Exchange Resins PWA5 A530E PGW6002E 0.53 0.3 0.65 A532E 0.33 Theoretical NTO exchange capacitya (meq/g) Observed NTO 1.16 ± 0.6 1.17 ± 0.5 1.06 ± 0.5 1.26± 0.6 exchange capacityb (meq/g) Observed NTO 0.85 ± 0.9 0.25 ± 0.2 0.58 ± 0.3 0.21 ± 0.1 desorptionb (meq/g) Difference 166 360 98 349 between Theoretical and Observed (%) a See Appendix II – A, b See Appendix II – B for supporting calculations. The results indicated that Purolite A532E, a gel, perchlorate selective resin had the greatest NTO exchange capacity of the resins investigated. However, it is difficult to discern how resin material composition or ion selectivity affect NTO exchange capacity based on the relative similarity of exchange capacities and small sample set. Regarding regeneration, all resins tested were capable of releasing NTO back into solution after being exposed to 25%-30% brine. However, the perchlorate selective resins (Purolite A530E and Purolite A532E) demonstrate a significant difference in desorption results compared to the nitrate selective resins. Again, it is difficult to discern the effects of resin material composition and ion selectivity. Further studies will need be replicated 16 in order to explore how resin material composition and ion selectivity affect exchange and regeneration performance. It should be noted that the analysis did not account for the effect of filters on NTO concentrations in calculating the resin exchange capacity. Additional details about the reasoning for excluding filter effects is included in Appendix II – B. All four resins were capable of more NTO than was expected: the observed exchange capacities are greater than the theoretical exchange capacity as shown, in general, it is difficult to make conclusions about the exchange capacities as these results as the studies were not replicated, thus the exchange values shown from these studies may not be representative of the resins’ exchange capacity. Future studies will include replicate studies in order to clarify the resins’ capacity to exchange NTO with chloride. Another observation to note is of the range in differences between the theoretical exchange capacity and observed exchange capacity across resins - 98% – 360 % difference, which is discussed in more detail in the Discussion section. 3.3 Complex Solution Ion Exchange and Regeneration Study Results The purpose of the complex solution study was to examine how NTO exchange compares with nitrate, a competing anion, in solution. The complex solution study indicated that Amberlite PWA5 resin is capable of exchanging NTO in the presence of nitrate in solution, as shown in Table 5. Amberlite PWA5 is a nitrate selective resin. 17 Table 5: NTO Exchange by Amberlite PWA5 Resin with Nitrate Present NTO = 300 NTO = 300 NTO = 300 NTO = 300 NTO = 300 mg/L, mg/L, mg/L, mg/L, mg/L, NO3 = 0 NO3 = 1 NO3 = 10 NO3 = 100 NO3- = 1000 mg/L mg/L mg/L mg/L mg/L Observed 1.4 1.3 1.2 1.0 0.8 NTO exchange capacity (meq/g) Observed 0.05 INC INC INC 0.69 NTO desorptionb (meq/g) INC = Results were inconclusive As expected, the NTO exchange capacity decreases as the nitrate concentration increases in solution. With no nitrate exposed to the resin, the NTO exchange capacity started at 1.4 meq/g and decreased to 0.8 meq/g as the nitrate concentration gradually increased to 1000 mg/L NO3- (Table 5). This result may be attributable to the fact that the resin was nitrate selective, and with an increasing amount of nitrate in solution there was a greater amount of nitrate to be exchanged. However, the regeneration results associated with the complex solution experiment were inconclusive, as the resulting NTO concentrations after regeneration for 24 hours were far greater than the expected 300 mg/L NTO initially exposed to the resins. Errors associated with the Hach spectrophotometer measurement might be associated with the particular result, but it is unlikely that NTO was produced after the regeneration process occurred, thus these results were considered inconclusive. From the perspective of nitrate exchange, the results shown in Table 6 indicated that the resin exchange capacity for nitrate in the presence of NTO in solution is lower compared to resin exchange capacity for nitrate when the solution was exclusively 18 nitrate. This result is expected, as competing ions in solution have the tendency to reduce exchange capacity for all ions in solution to some degree. However, this suggests that the resins have an affinity for NTO that may be significant compared to nitrate. Supporting calculations are shown in Appendix II - C. Table 6: Nitrate Exchange by Amberlite PWA5 with NTO Present Solution: NTO + NO3Solution: NO3Observed NO3 exchange 0.23± 0.1 0.72± 0.6 capacity (meq/g) Observed NO3- desorption 0.21± 0.2 0.36± 0.38 capacity (meq/g) As with the analysis of NTO concentrations, the effect of filters on nitrate concentrations during sampling were not accounted for in calculating the nitrate exchange capacity. Additionally, replication experiments were not conducted, thus the exchange values shown from the complex solution studies may not be representative of the resin’s exchange capacity for nitrate. 4 DISCUSSION 4.1 Variability of Exchange Capacity Results Differences between the theoretical exchange capacity and observed exchange capacity of NTO ranged between 98% - 360% across all resins tested. The high percent difference in theoretical exchange capacity and observed exchange capacity is likely attributable to errors in the analysis of resin exchange capacity. However, the wide range of exchange capacities may also be attributable to the variety the resins used, including differences in resin material and selectivity. The resin material composition can affect the exchange capacity performance. Gel resins are essentially homogenous solid beads made of tightly spaced functional groups 19 on the bead, whereas macroporous beads are made of networks of functional pores [18] (Figure 6). Gel resins have the tendency to have a greater capacity and better regeneration efficiency than macroporous resins, but macroporous resins have a highly cross-linked structure, making them preferred in special applications [19]. A study investigating the interparticle diffusion (kinetics) associated with a gel resin versus a macroporous resin in exchanging anionic pentachlorophenol (PCP-) found that gel resins were quicker to facilitate exchange but were less receptive to improving exchange performance with an increase in chloride ion concentration in solution compared to the macroporous resin [20]. This result may be attributable to non-continuous structure of macroporous resins, which tends to create slower pore diffusion[20]. Figure 6 – Gel type resins and macroporous resins structure [20]. The results of the PCP- study are informative to the results of this work (PCP- and NTO are similar in that they are both anionic, organic compounds) as it suggests that the gel resins should exchange greater concentrations of NTO for chloride than the macroporous resins. While the results of this work do not strongly suggest that the gel resins had a greater exchange of NTO than the macroporous resins, the PCP- study does validate that 20 it is reasonable to see differences between the exchange performances between the resins examined. The ion selectivity of the resins also plays a role in the differences observed in exchange capacities among resins. The resins evaluated were either nitrate or perchlorate selective, which would have affected the exchange of NTO as the “differences between the hydrated radius of NTO compared to the hydrated radii of nitrate and perchlorate to which the resins were designed to” create inherent variability between the performance of nitrate and perchlorate resins in exchanging NTO [7]. 4.2 Batch versus Column Study Considerations The studies conducted for this work were done in batch systems, with a setup similar to what is shown in Figure 5. Experiments were conducted by batch study for simplicity. Figure 7 – IX Batch And Column Setup IX batch studies produce exchange capacity results that are dependent on the exchange observed during experimentation, whereas IX column studies would produce exchange capacity results that are a function of the amount of resin material and flow rate. While 21 batch studies were sufficient enough to examine whether or not the resins were capable of exchanging NTO for chloride, a column setup would be necessary in order to provide information needed for building an industrial scale IX treatment system. Future studies will investigate the same conditions as done in this report (i.e. exchange, regeneration, kinetics, and complex solution) but with a column setup (Figure 7). 4.3 Industrial Scale IX System Consideration The exchange capacity values found in this study may not be representative of the resin’s performance, but the trends found in the ion exchange, regeneration, kinetics, and complex solution results are reasonable enough to assume that further investigation will produce the parameters necessary to design a proper IX treatment system that can treat NTO from wastewater streams. It is expected that on an industrial scale, the NTO exchange capacity of the resin will be lower than what laboratory results may indicate, as wastewater streams are much larger and contain heavier loads of cations and anions. Additionally, the concentration of brine required to regenerate the resins will likely not need to be as concentrated as the 25% - 30% brine focused in this study, as many resin manufacturers specify a 10-15% brine range as sufficient for regeneration purposes. Alongside the additional ion loading, there will also be other munitions present in the wastewater stream, such as DNAN and NQ, which could impede the exchange of NTO by the anion exchange resins. While DNAN and NQ are not receptive to IX (both are not ions at a neutral pH), they could still impede the resin exchange performance by interfering with the pathway between NTO and the resin exchange sites. IX would be placed at the end of a treatment train as IX benefits from upstream pretreatment (which includes coagulation, filtration of suspended 22 solids, application of anti-scalant chemicals, and/or softening) in order to extend the IX replacement period and improve resin exchange performance [21]. 5 CONCLUSION Overall, the results of the studies demonstrate that NTO can be removed from wastewater by nitrate and perchlorate anion exchange resins. Removal of NTO by an anion exchange resin does meet a maximum exchange capacity during a 24 hour period, and NTO can still be exchanged with competing ions in solution. These results provide new insight into alternative industrial wastewater treatment technologies that could be used at munitions manufacturing and LAP facilities. Future work will replicate the ones included in this work, accounting for methodological errors made in order to validate the exchange capacities observed in this study. In addition, further work will be done to translate the laboratory results to an industrial scale by transitioning from IX batch to column setup study. Simultaneously conducted but outside the scope of this work, studies are also being done on the biotic degradation of NTO, which would further explore the possibility of treating NTO sustainably and completely. 23 6 APPENDICES 6.1 Appendix I: Additional Standard Curve Equations Table 7: Standard Curve Equations for NTO and NaCl Solutions Standard Curve Equation R2 value NTO – 0 % NaCl CNTO = 0.0005(AreaLC) + 9.7332 0.9989 NTO – 12 % NaCl CNTO = 0.0005(AreaLC) + 6.0532 0.998 NTO – 15 % NaCl CNTO = 0.0005(AreaLC) – 0.5947 0.9995 NTO – 18 % NaCl CNTO = 0.0005(AreaLC) – 8.5417 0.9955 NTO – 25 % NaCl CNTO = 0.0005(AreaLC) – 1.3888 1 NTO – 30% NaCl CNTO = 0.0002(AreaLC) + 0.2403 0.995 6.2 Appendix II: Exchange Capacities 6.2.1 Appendix II - A: Theoretical Exchange Capacities The theoretical exchange capacity was calculated by multiplying the total exchange capacity of each resin, average moisture capacity, and resin density (same sources as indicated in Table 1). Resin densities were calculated by using the specific gravity of each resin as listed by sources. The exchange capacity was an average of exchange capacities associated with different initial concentrations of NTO. 24 Table 8: Theoretical Exchange Capacity of Anion Exchange Resins PWA5 A530E PGW6002E A532E Total 1 0.6 1.65 0.85 Exchange Capacity (eq/L) Average 55 52 42.5 40.5 moisture (%) Resin density 1036.88 1036.88 1086.73 1036.88 3 (kg/m ) Theoretical 0.53 0.30 0.65 0.33 Exchange Capacity (meq/g) 6.2.2 Appendix II - B: Ion Exchange and Regeneration Exchange Capacities For all studies, Equation 5 was used to determine exchange capacities (meq/g): Equation 5: NTO Exchange Calculation ∆ 𝑚𝑚𝑚𝑚 1 𝑚𝑚𝑚𝑚𝑚𝑚 1𝑒𝑒𝑒𝑒 1 𝐿𝐿 1 × 30 𝑚𝑚𝑚𝑚 × × × × = 𝑚𝑚𝑚𝑚𝑚𝑚/𝑔𝑔 𝐿𝐿 130 𝑔𝑔 𝑁𝑁𝑁𝑁𝑁𝑁 𝑚𝑚𝑚𝑚𝑚𝑚 𝑁𝑁𝑁𝑁𝑁𝑁 1000 𝑚𝑚𝑚𝑚 0.05𝑔𝑔 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 Exchange capacity was determined by comparing the initial concentration of NTO in solution to the concentration of NTO in solution after being exposed to a resin for 24 hours. It was assumed that the initial concentration of NTO in stock solution would substitute the time zero sample, as a time zero sample was not taken. Regeneration exchange capacity was determined by comparing the initial concentration of NTO in solution to the concentration of NTO from the ion exchange study to the concentration of NTO after resins were exposed to brine solution for 24 hours. Previous studies indicated approximately 50% NTO loss associated with using a filter during sampling. When accounting for these effects in analyzing the data (multiplying results by two), the concentration results were higher than the known NTO concentration exposed to the resin, thus the effects of the filter were not included in 25 evaluating the exchange capacity of the resins. For example, the Amberlite PWA5 resin was exposed to approximately 500 mg/L NTO. Samples of the original NTO solution used did not use a filter. After 24 hours, a sample was taken with a filter, and the resulting amount of NTO not exchanged by the resin was approximately 400 mg/L. Accounting for the expected 50% NTO loss by filter would indicate that after 24 hours, the resulting NTO in solution would be approximately 800 mg/L, which is beyond the expected and measured initial NTO concentration. Several results demonstrate this behavior when the effect of the filter is accounted for, so for the simplicity of analysis, it was assumed that the effects of the filter were negligible. Future studies will re-evaluate the concentration losses associated with the filter and account for them in future analysis. 6.2.3 Appendix II - C: Complex Solution Ion Exchange and Regeneration Exchange Capacities Equation 6 was used to determine exchange capacities (meq/g): Equation 6: NO3- Exchange Calculation ∆ 𝑚𝑚𝑚𝑚 1 𝑚𝑚𝑚𝑚𝑚𝑚 1𝑒𝑒𝑒𝑒 1 𝐿𝐿 1 × 30 𝑚𝑚𝑚𝑚 × × × × = 𝑚𝑚𝑚𝑚𝑚𝑚/𝑔𝑔 𝐿𝐿 62 𝑔𝑔 𝑁𝑁𝑁𝑁𝑁𝑁 𝑚𝑚𝑚𝑚𝑚𝑚 𝑁𝑁𝑁𝑁𝑁𝑁 1000 𝑚𝑚𝑚𝑚 0.05𝑔𝑔 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 Nitrate exchange capacity was determined by comparing the time zero sample of nitrate in solution to the concentration of nitrate in solution after being exposed to a resin for 24 hours. The time zero sample occurred after exposing the resin to the nitrate solution. Regeneration exchange capacity was determined by comparing the time zero sample of nitrate in solution to the concentration of nitrate from the ion exchange study to the concentration of nitrate after resins were exposed to brine solution for 24 hours. Results indicated approximately 62% nitrate loss associated with using the filter during sampling events. As noted for the analysis of NTO concentrations, accounting for losses 26 associated with the filter in analysis led to results that were unexpected, thus the effects of the filter were considered negligible. 27 7 ACKNOWLEDGEMENTS The author would like to thank Dr. Jennifer Weidhaas, Saru Poudel-Acharya, and the entire University of Utah Department of Civil and Environmental Engineering. This project was funded by the University of Utah Office of Undergraduate Research. 28 8 [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] REFERENCES S. C. Chew, M. Tennant, N. Mai, D. McAteer, and J. F. Pons, “Practical Remediation of 3-Nitro-1,2,4-triazol-5-one Wastewater,” Propellants, Explos. Pyrotech., vol. 43, no. 2, pp. 198–202, 2018. Matthew W. Smith and Matthew D. Cliff, “NTO-Based Explosive Formulations: A Technology Review.” L. Le Campion, A. Vandais, and J. Ouazzani, “Microbial remediation of NTO in aqueous industrial wastes.” D. Felt et al., “Evaluation of Treatment Technologies for Wastewater from Insensitive Munitions Production; Phase 1: Technology Down-Selection. Report: ERDC/EL TR-13-20.,” US Army Eng. Res. Dev. Center, Environ. Lab. ERDC/EL TR-13-20, no. November, 2013. S. Panja, D. Sarkar, and R. Datta, “Vetiver grass (Chrysopogon zizanioides) is capable of removing insensitive high explosives from munition industry wastewater,” Chemosphere, vol. 209, pp. 920–927, Oct. 2018. T. Richard and J. Weidhaas, “Biodegradation of IMX-101 explosive formulation constituents: 2,4-Dinitroanisole (DNAN), 3-nitro-1,2,4-triazol-5-one (NTO), and nitroguanidine,” J. Hazard. Mater., vol. 280, pp. 372–379, Sep. 2014. J. Weidhaas, S. Acharya P, D. Tran, V. Gupta, and R. Goel, “Ion Exchange and Biological Treatment of NTO Wastewater,” in JANNAF Conference, no. 1, p. 7. Rohm and Haas, “Ion Exchange for Dummies.” “Ion Exchange Introduction.” [Online]. Available: http://dardel.info/IX/IX_Intro.html. [Accessed: 20-Apr-2019]. P. S. Bä Uerlein, T. L. Ter Laak, R. C. H. M. Hofman-Caris, P. De Voogt, and S. T. J. Droge, “Removal of charged micropollutants from water by ion-exchange polymers e Effects of competing electrolytes,” 2012. E. Atikovic, M. T. Suidan, and S. W. Maloney, “Anaerobic treatment of army ammunition production wastewater containing perchlorate and RDX,” 2008. W. E. Platten, D. Bailey, M. T. Suidan, and S. W. Maloney, “Biological transformation pathways of 2,4-dinitro anisole and N-methyl paranitro aniline in anaerobic fluidized-bed bioreactors,” Chemosphere, vol. 81, no. 9, pp. 1131–1136, Nov. 2010. D. L. Freedman and K. W. Sutherland, “Biodegradation of hexahydro-1,3,5trinitro-1,3,5-triazine (RDX) under nitrate-reducing conditions,” Water Sci. Technol., vol. 38, no. 7, pp. 33–40, Jan. 1998. “AMBERLITETM PWA5 Resin Drinking Water Grade.” “Purolite® A530E.” [Online]. Available: https://www.purolite.com/en/product?normalizedName=a530e&simple=true. [Accessed: 11-Apr-2019]. “Purolite® PGW6002E.” [Online]. Available: https://www.purolite.com/product/pgw6002e. [Accessed: 11-Apr-2019]. “Purolite® A532E.” [Online]. Available: https://www.purolite.com/product/a532e. [Accessed: 11-Apr-2019]. “Intraparticle Diffusion during Selective Sorption of Trace Contaminants: The Effect of Gel versus Macroporous Morphology,” 2000. 29 [19] “Dow Answer Center.” [Online]. Available: https://dowac.custhelp.com/app/answers/detail/a_id/437/~/dow-ion-exchangeresins---gel-vs.-macroporous-resin. [Accessed: 17-Apr-2019]. [20] E. Van de Steene, J. De Clercq, and J. W. Thybaut, “Ion-exchange resin catalyzed transesterification of ethyl acetate with methanol: Gel versus macroporous resins,” Chem. Eng. J., vol. 242, pp. 170–179, Apr. 2014. [21] “Ion Exchange vs. Reverse Osmosis: Choosing the Best Treatment System for Your Needs.” [Online]. Available: https://www.samcotech.com/ion-exchange-vsreverse-osmosis-choosing-best-treatment-system-needs/. [Accessed: 23-Apr2019].