Interactions of Polymers and Energetic Materials

Trace explosive detection is the primary way for quick and easy sampling of various surfaces in a check-point environment, e.g. an airport. The swabs used for commercial explosive detectors, such as ion mobility spectrometers, are made of various materials. The difficulty in collection and analysis of explosive traces is that the swabs must be effective at adsorbing as well as desorbing materials, i.e. pickup from the surface and release into the detection device. This dichotomy results in a tradeoff for development of new swabs. Generally, desorption is considered to be the more desirable property; therefore, Teflon is one choice for commercial swabs. It would be ideal to develop a swab with both enhanced adsorption and desorption. One way to accomplish this is to apply an electrostatic charge to commercial swabs. This enhances their attraction to explosive particles, but once the swab is placed in the inlet of commercial detection instrument, the charge is dissipated, and desorption of the particles into the instrument proceeds as usual. Methods of generating electrostatically charge swabs was determined; triboelectric charging vs corona charging was compared examining magnitude of the charge, reproducibility and stability, and effects of humidity. The magnitude of charge necessary for enhanced collection of particles was evaluated using an electrostatic voltmeter to measure charge and various means to measure particle pickup. Corona charging was determined to be more effective. Enhancement of collection was judged by comparing results of corona charging swabs to those achieved by contact swabbing. Two variables were examined: the analyte and the substrate from which the analyte is removed. The swab material was Nomex. In each case but three, collection of an analyte by an electrostatically enhanced swab outperformed the traditional contact swabbing. Evaluation was determined by a rigorous quantification by mass spectrometry of the analyte picked up by the swab and the analyte remaining on the substrate after swabbing. When analytical protocol was not amenable to a particular analyte or substrate a commercial explosive trace detection instrument was used. It was found that the substrate morphology played a bigger role in pickup of analyte than the particular analyte. In order to eliminate biological warfare agents, both heat and halides are used. Ideally, these agents would be destroyed without dispersing them. The approach to create a polymeric-sprayable matrix would allow dispersion of an iodine-producing pyrotechnic, without dispersing the biological weapon, and when initiated would produce both heat and iodine gas. This matrix will provide iodine vapor and a longlasting flame, not an explosion, to control dispersion of the threat. A two-part foam was formulated based on polyurethane chemistry, i.e. a diisocyanate combined with polyol to produce a urethane linkage. Each component of the foam (e.g. isocyanate, polyol, catalyst, blowing agent, surfactant) was experimentally adjusted to achieve the best foam based on expansion, structural integrity, and cell uniformity. Since the polyol is the most adjustable component in the foam, an investigation of commercial and synthesized energetic polyols was performed. The structures of the energetic polyols were verified by LC-MS and FTIR and characterized for heat flow by DSC. Once the structures of the energetic polyols had been proven, it was formulated into a polyurethane foam which was characterized for heat of decomposition, by SDT, for heat of combustion by bomb calorimetry, and structurally by FTIR. Documenting heat flow with SDT helped to determine that the structural modification increased heat release and lowered ignition temperature compared to the standard polyurethane foam. The formulated polyurethane foam was then tested for expansion against increased solids loading. When optimal solids loading was determined (>70%), the pyrotechnic foam was ignited in a bomb calorimeter. The heat released and iodine production was quantified.

viii LIST OF TABLES      [3]. A dichotomy in explosive trace detection is that swabs must be effective at sorption as well as de-sorption, i.e. sorption to the surface and release into the detection device.
Generally, de-sorption is considered to be the more desirable property; therefore, polytetrafluoroethylene (PTFE) known as Teflon, is one choice for commercial swabs [2]. Traditionally, swabbing is by direct contact. The contact swabbing of hands, headdresses, and medical appliances can increase the amount of contaminants on the swab and can be personally invasive. The ideal swabs would both enhanced sorption and de-sorption and require no contact with the subject or object.
This study considers application of electrostatic charge to commercial swabs.
This can be accomplished three ways: triboelectric charging (friction), induction charging, or conduction. The classic conductive charging method uses a Van de Graaf 4 generator. The head of the Van de Graaf generator is negatively charged, and when a neutral object comes into contact with it, the charge is transferred [4]. This method was initially examined, but the charge was non-directional and short-lived on surfaces.
Thus, only triboelectric and corona charging were further investigated.
The static charge created on the two surfaces can be positive or negative, depending on their relative position in the triboelectric series [7]. The triboelectric series lists materials in order of their preference to obtain a charge of relative magnitude in  The further away two materials exist on the triboelectric series, the higher the resulting magnitude of charge on the two materials. For example, PTFE is highly negative on the triboelectric series, and friction between it and a highly positive material in the series should result in two oppositely charged materials [5][6]9]. This method of charging is termed "triboelectric charging." This method might also result in transfer of contaminants. Inductively charging a swab can be achieved without formed by exposing a material to polarized electric fields at high temperature [11][12][13][14].
In theory, any material with a measurable relative permittivity should be subject to electrostatic enhancement [15][16]. In this work, the potential for noncontact electrostatic swabbing was investigated and benchmarked to performance of standard contact swabs. verified by FTIR using attenuated total reflectance (ATR) was also used.

Scanning Electron Microscopy (SEM)
Scanning electron micrographs were obtained at 20 kV in secondary ion mode on a JEOL 5900 SEM of some of the substrates and swabs. Secondary ion mode was used because it specifically focused on the surface morphology.

Analytes
Initially, commercial sugar and sodium chloride were used as surrogates for organic explosives and inorganic energetics, respectively. These were sieved to  A humidity chamber was built using a humidistat and glove box. Controlled humidity environments were achieved with saturated salt solutions [17]. Experiments at 0% humidity used dry compressed air.

Dry Transfer
Trace quantities of solid analyte were placed on surfaces as solutions of known concentrations via micro-syringe and solvent allowed to evaporate. This resulted in an unrealistic distribution of analyte on the surface (e.g. coffee rings); therefore, a dry transfer method was developed [18]. The solution of analyte was placed by Eppendorf pipette (2-20 μL) onto a strip of Bytac®. The solvent was allowed to evaporate from the Bytac, and the resulting residue on the Bytac® was transferred to the desired substrate by rubbing the Bytac over the substrate in one direction. Curves were analyzed and linearly correlated between the ratio of internal standard and the analyte which was weighted (1/x 2 ) by Thermo Xcalibur Quan Browser software version 3.0.63. An example of a standard curve can be seen in Figure 1 Dynamic range for PETN analysis was from 10 to 2500 ng/mL; for TNT, 25 to 5000 ng/mL; and for RDX, from 10 to 2500 ng/mL.

Bulk Measurements
As a proof of concept, electrostatic pick up of visible amounts of surrogate analytes, sucrose and sodium chloride, was performed ( Figure 1

Method for charging swabs
To charge swabs inductively, a pinner was fixed over an unheated hot plate base which had been covered with a perforated metal (zinc coated steel) grounding plate ( Figure 1.8). The grounding plate had a screw in one corner and a wire which was fixed to the screw of a traditional electrical outlet (A schematic is shown in Figure   1.9). The contact was tested with a multi-meter to verify proper grounding. Since the use of inductive charging relies on the attraction of a charge through the swab to a base, various bases were examined (Table 1.5). Tests were performed charging at -20 kV and 2 inches from the swab for 5 seconds at 40% RH. Voltages were measured using a 3M 718 electrostatic voltmeter, and a static dissipater was used to eliminate excess charge before each experiment. Results, shown in Table 1

Humidity Measurements
Humidity affects the generation of electrostatic charges both on the molecular and macroscopic level [19]. In electrostatic precipitators, it has been shown that collection efficiency of small particles (<50 μm) increases with increasing humidity, and that negative corona discharge is less sensitive to humidity than positive [20]. The effects of humidity were examined on three swabs. They were held in air over a 3-min time period (Figures 1.10, 1.11, 1.12). In the 0% RH-30% RH regime there was little change in the initial magnitude; the decay was less than 1 kV; however, in the 50% RH and 80% RH regime, a significant drop in initial magnitude was observed. This was consistent with literature reports that at low levels of humidity, the charges are inserted directly into the bulk volume, whereas if moisture can accumulate on the surface of the material, the charge decays slowly into the bulk [21].

Swabbing and Swab Preparation
For contact swabs, they were physically rubbed against the substrate with the experimenter attempting to apply ~7 N of force [22][23]. Electrostatically enhanced swabs were charged by the inductive charging method and then passed along the substrate from a distance of ~10 mm.

Charging and storage of swabs in stacked and unstacked configuration
Commercially purchased swabs are packaged in multi-swab packages. 20 In another experiment, tests were performed (trial 1 at 40% RH and Trial 2 at 30% RH) in which 10 Nomex or 10 PTFE swabs were charged simultaneously at -20 kV for 5 seconds in an unstacked configuration as shown in Figure    22

Quantification of Analytes
The remainder of this study examined pickup efficiency of electrostatic swabs compared to contact swabs. Substrates examined are shown in (Figure 1.16  The experiment to examine pickup from each substrate was conducted in three parts. First, to determine the extraction efficiency from each substrate, three substrates were directly spiked with 500 ng of analyte in solvent; the solvent was allowed to evaporate; the substrates were extracted with another solvent; the solvent was adjusted to proper volume; and a portion was analyzed by LC/MS. The result indicated the amount of the original 500 ng of explosive that could be recovered from the surface of a given substrate. An example of this using polycarbonate (PC) as substrate S1 is shown in top 3 lines Table 1.10.
Then six samples of Bytac® were spiked with 500 ng of the analyte and rubbed against substrate the first substrate S2 and both S1 and S2 were analyzed to the dry transfer efficiency for a given substrate (middle section of Table 1.10). Three of these S2 and the three S1 against which they were rubbed (dry transferred) and were 23 extracted with solvent; the solvent was adjusted to proper volume, and was analyzed by LC/MS. The result indicated the amount of explosive that could possibly be transferred from S2 to the swab. That amount plus that remaining on S1 ideally would add to 500 ng (1, middle section). Then, six more Bytac® (S1) were directly spiked with 500 ng of analyte, dry transferred to S2, then three of the S2 were swabbed by contact, i.e. directly rubbing the Nomex swab against S2 (~7 N force) and three of the S2 were swabbed using the electrostatically enhanced Nomex swabs.
Then, these six swabs, S2, and S1 were extracted with solvent; the solvent was adjusted to proper volume; and a portion was analyzed by LC/MS. For any given set, the amount of explosive left on S1 and S2 and found on the swab should have added to 500 ng. The amount of explosive found on the three electrostatically enhanced swabs versus the three contact swabs is shown in Table 1 Table 1.10, 351 ng were found on the swab out of a total recovery of 475 ng. This translates to a 74% recovery.

Polycarbonate
The results shown in Table 1.10 for both contact and electrostatic swabs are below 100 ng, except for one outlier (3 rd electrostatic swab). These results suggest analyte on polycarbonate is difficult to sorb onto a swab, whether contact or electrostatic. This could be related to the smoothness of the polycarbonate surface.

Ohio Travel Bag Zipper
In comparison to some of the other substrates which are flat surfaces, the Ohio zipper is a small pull tab with many grooves (Figure 1.17). According to the manufacturer, the pull tab is made of a polyurethane resin. The results obtained in Table 1 This is to be expected as a residue which was direct deposited via solvent can settle in the grooves of a substrate as solvent evaporates. This is in opposition to dry transfer where the analyte tends to adhere to the more exposed surface. In the case of TNT from aluminum foil, 41

Packing Tape
The objective in determining pickup from polypropylene packing tape was to examine the non-sticky side. It was necessary to remove the adhesive backing for subsequent extraction of the analyte. However, no amount of extraction completely removed the adhesive. FTIR analysis (Figure 1.18) revealed the tape was made of polypropylene. Therefore, polypropylene was used as substrate S2. Data in Table 1.29 shows PETN deposited on polypropylene resulted in recoveries <100%.  No further attempts to quantify on LC/MS with these substrates occurred in this study.

Explosive Trace Detectors (ETD)
In this section the feasibility of using electrostatic non-contact swabbing methods with airport ETDs is verified. Shown in Figures 1.19   In Table 1.30, the dry transfer method was used to deposit 100 ng of PETN, 100 ng TNT, or 500 ng AN on the various substrates listed in the substrate 2 column.
These substrates were swabbed with either an electrostatically enhanced swab or a contact swab; the swab was presented to the ETD, and the magnitude of the response was recorded. The results show electrostatically enhanced swabs achieved higher response magnitudes than contact swabs. In the cases of ballistic nylon, polycarbonate (PC), PTFE, and ABS neither electrostatic nor contact swabbing was able to elicit responses from the instrument at this concentration. This trend agrees with LC/MS quantification showing that neither electrostatic nor contact swabbing had a greatly enhanced performance. In the vinyl and Nylon 1000 experiments, electrostatic swabs retained a higher amount of analyte than contact. Results agree with literature that transfer efficiency is poor to hard plastic [24].

C-4 Fingerprints: Comparing positive and negative voltages
In order to determine if the sign of the voltage had an effect on sorption of explosive residue, a simple test was completed (Table 1.31). With increasing magnitude, whether positive or negative, response magnitude increases, the only outlier being the first test of +4.78 kV where the response magnitude is the highest.
Two swabs (fiberglass coated Teflon) were stacked on the grounding plate. The swabs were charged for 5 seconds at -20 kV from 2 inches away from the pinner at 40% RH.
Single C-4 fingerprints were made with index fingers of various participants on a cellulose substrate (filter paper). Charges were recorded and swabs approached fingerprints from 10 mm above the substrate. Swabs were tested in the Morpho Itemiser DX ETD. Additionally, controls were performed where charged swabs went directly into the inlet to insure no false alarms came from charging. These results verified that both positive and negative charging results in a response from the IMS. 46 CONCLUSION Commercial swabs were charged using the inductive method. This charging method was optimized for the surface on which charging occurred and length of charging time (5 sec). Several commercial swab types were tested in humidity ranging 0 to 80% RH. As humidity increased charge imparted to the swabs decreased. The negatively charged swabbing material still developed enough charge to enhance particle pickup. Cotton swabs charged positively, and rapidly lost this charge. Storage and stacking methods were tested to determine if swabs could be packaged charged; this did not seem to be the case.

INTRODUCTION
The threat of biological weapons, specifically spores of Bacillus anthracis species, are of concern to those charged with homeland security. Although the exact kill mechanism is unknown, current research efforts suggest that heat and iodine gases can act as a biocide causing DNA damage and increasing the kill-rate in spores [3][4].
Previous work examined the heat and iodine outputs of various fuels and oxides of iodine mixtures [5]. For controlled application and dispersion of the biocidal formulations in the field, these fuel/oxidizer mixtures require a binder. For a quick and thorough application of the biocidal formulation, a sprayable binder was considered best. The properties desired in a sprayable binder would include rapid curing (less than 10 seconds) with significant expansion and acceptable mechanical properties so that high solids loading could be achieved. It was determined that polyurethane foams met these needs. Because polyurethane formations are based on combining a diisocyanate with a polyol to produce a urethane linkage, (Figure 2.1) a number of modifications can be applied by changing one of the monomers [6]. Functionalizing polymers with nitro or azide groups is the usual approach to creating energetic binders [7][8][9]. Polyurethane foams were exclusively investigated in this study.

Materials
All chemical reagents and analytical grade solvents were purchased from, Acros Organics, Alpha Aesar, Fisher Scientific, TCI, or STREM and used without further purification. Particle size of the calcium iodate was 70 to 150 μm; and the aluminum, from Obron, was 23 μm. To lower the surface tension during polymer preparation, promote uniform cell structure, and increase expansion, surfactant (Triton TM GR-7M) provided by DOW Chemical was used at 1 wt % solution in pentane.

Fourier Transform-Infared Spectrometer (FTIR)
Infrared spectra were measured with a Thermo Scientific Nicolet 6700 spectrometer equipped with a Smart iTR diamond ATR. FTIR spectra were recorded at ambient temperature. Background and spectra were collected in ranges of 4000-650 cm -1 .

Thermal Measurements
Melting points, decomposition temperatures, and enthalpies were determined using a TA Q100 for differential scanning calorimetry (DSC), calibrated against an indium standard, heating at 20 C min -1 under nitrogen flow of 50 mL/min. Hermetic aluminum pans were used for monomers, and glass sealed capillaries, for polymers.
Polymer decomposition was monitored with a TA Q600 Simultaneous DSC/TGA (SDT) instrument; samples were held in open alumina pans, with heating rate 20 C min -1 and nitrogen flow of 300 mL min -1 .

Liquid Chromatography/ Mass Spectrometer (LCMS)
To determine mass spectra, each compound was dissolved in methanol to make a      Increasing concentration of the catalyst increased the rate of foam formation; a near instantaneous reaction was desired [17][18]. If applicable the specially synthesized monomer, and a small amount of acetone was used to adjust viscosity. For the standard TEOA-TDI synthesis, Part B was added to Part A, for synthesis involving the special monomers, Part A was added to Part B'. The reaction took 5 seconds, increasing in temperature as it cured to form the standard polyurethane (TEOA-TDI); it exhibited significant expansion (9X) and hardening to withstand stimulation with a glass stir rod.
The exothermic cure of polyurethane foams can reach temperatures at high as 140-170 °C which result in 'scorching' of the polymer. Scorching weakens the urethane linkages so that they no longer withstand the mechanical stress associated with high solids loading [6]. The TEOA-TDI foam was white in color with no visible scorching (Figure 2.7). This synthetic method was altered using special energetic monomers, starting at 50:50 mole ratios, and varying until foam had a desirable structure and expansion factor.  Starting at 50:50 mole ratio of TETRA to TEOA, the amount of TETRA 2 was reduced until the TETRA to TEOA ratio was 0.4. Even at this low ratio, scorching was observed in the center of the polymer mass. Lowering catalyst to 50 µL didn't eliminate scorching.

Foams with Pyrotechnic
To examine how significant amounts of solids loading affected polymer formation or expansion, the following experiment was performed. Part A was placed in a centrifuge tube and stirred. Part B, which consisted of TEOA and TEA, was added to Part A. Immediately after the reaction began, discernable by yellow coloration, and before bubbling commenced, the pyrotechnic was added, an amount equivalent to 73% solids loading. The pyrotechnic formulation was a 90/10 Ca(IO3)2/Al mixture, 73% solids loading is equivalent to 4.84 g pyrotechnic per 1.80g foam. The reaction mixture was stirred once before it began to expand; the reaction was complete in 20 seconds or less.
The product was porous, grey in color, and exhibited similar expansion (8X) to the foam without solids.

Scanning Electron Microscope (SEM)/ Energy Dispersive Spectroscopy (EDS)
Pyrotechnic-loaded polyurethane foam samples were dried in a vacuum oven and transported in falcon tubes. Small slivers were cut with razor blade from the top and bottom of the expanded pyrotechnic-loaded foam. Samples were placed on the sample holder examined under low vacuum (25 Pa) in backscatter mode at a voltage of 20 kV.

Powdered Foams with Pyrotechnic
For calorimetry, foams were ground in a coffee grinder and sieved to particle sizes of 150-300 µm. They were mixed in a Resodyn Acoustic Mixer with pyrotechnic.

Bomb Calorimetry/UV-VIS
Bomb Calorimetry was performed using a Parr 6200 Isoperibol Bomb Calorimeter. The calorimeter was calibrated (10 trials excess Iwhich solubilized I2 and transformed it to I 3-(λ=353 nm) [19]. An example of a calibration curve for iodine quantification is shown in Figure 2.12. In Figure 2.13, the FTIR of each monomer used is shown. At around 3200 cm -1 the OH stretch is visible and broad for all cases. In spectra 1,2,3 and 5, the stretch for NO2 are seen at 1560-1510 cm -1 . In spectra 3 and 4, the N3 stretch is sharp at 2100 cm -1 .

LCMS
In order to confirm the synthesis of the monomers, LCMS was performed. Figure   2.14 shows the results from the mass spectrometer. TETRA (2)

Pyrotechnic Foams
The standard foam TEOA-TDI was used to test how much pyrotechnic powder and how uniformly this powder could be introduced to the foam. The pyrotechnic mixture was added to TEOA-TDI in amounts ranging from 30 wt% to 80 wt%. It was determined that up to 73 wt% powdered pyrotechnic could be added without significantly compromising expansion (expansion was 8 fold rather than 9). To determine the uniformity of distribution of the pyrotechnic, a top and bottom sliver of the pyrotechnic-loaded TEOA-TDI foam was examined by SEM/EDS. Distribution was less uniform than desirable.
The micrographs in Figure 2.20 show more pyrotechnic in the bottom slice than the top.  Table   2.1 were mixed with 70 wt% pyrotechnic mixture of 90/10 Ca(IO3)2/Al and the heats of combustion and iodine production determined by bomb calorimetry. TETRA(2)/TEOA-TDI was not chosen for further study because its charring suggested that mixing with a pyrotechnic might result in hazardous, unexpected ignition. To overcome the lack of uniform distribution of the pyrotechnic in the foams, for calorimetry only, the foams were powderized, and 70 wt% of the pyrotechnic (90/10 Ca(IO3)2/Al) was mixed in (Table   2.2). SDT of the pyrotechnic-loaded energetic and standard foams (Figure 2.22) show the pyrotechnic mixture releases heat earlier and over a longer range than without an organic matrix.  Within standard deviation the pyrotechnic-loaded energetic foams do not outperform the standard foam. This was not the case for the polymers when no pyrotechnic was present (Table 2.1). Apparently, the boost in energy and oxygen production provided by the pyrotechnic (70 wt% loading) overwhelms any addition boost available from the energetic monomer which makes up less than 10 wt% of the overall formulation.

CONCLUSIONS
This study sought to improve the energy output of polyurethane/pyrotechnic formulations by replacing part of the polyol TEOA with polyols containing nitro and/or azide groups. Monomers containing one or two of these functional groups were synthesized, but it was found that the presence of two nitro groups in the polyol, as in Compounds 2 and 5, resulted in a material so energetic that when combined with TDI, the polyurethane could not be produced without "scorching." It should be noted that polyurethane resins have been made from compounds 4 and 5, but in that report polymerization was allowed to progress slowly (overnight) [14]. In contrast, the diazo group, one nitro group, and one azo/one nitro were successfully inserted into a polyurethane. We believe TMNM 1 and AZONITRO 3 polyurethanes are reported here for the first time.
It was found that a pyrotechnic powder could be mixed with the polyurethane foams at levels up 73% solids with little negative impact on the foam expansion; it was only reduced from 9 to 8 fold. Furthermore, the pyrotechnic added could contain less aluminum since fuel was supplied by the polyurethane. Whereas at least 25% aluminum was required to produce ignition in a Ca(IO3)2 powder, in the presence any of the polyurethane foams the oxidizer/aluminum ratio could be raised to 90/10 and still be readily ignited. Reduction of aluminum resulted in increased output of molecular iodine.