Trace Metal Capture

Trace metal measurement techniques are developed in the Clean Energy Conversions lab and combustion flue gas is simulated by burning methane in air. Breakthrough and adsorption experiments are carried out to investigate sorbents and oxidation catalysts for trace metal (mercury, arsenic, and selenium) and carbon capture.

Topic 1: Flue Gas Simulations
Trace metal reaction chemistry is experimentally investigated in the Clean Air Conversion lab. Combustion flue gas is simulated by burning methane in air.
Figure 1. Photographs from left to right are, flow reactor, EI-QMS, entrained-flow reactor. Images are duplicated with sketches. Only a sketch of the packed-bed reactor is pictured
Figure 1. Photographs from left to right are, flow reactor, EI-QMS, entrained-flow reactor. Images are duplicated with sketches. Only a sketch of the packed-bed reactor is pictured
Elemental mercury vapor is introduced from a mercury calibration system (Cavkit) precombustion. Arsenic and selenium oxides and chlorides are generated from a similar instrument, in which salts are gasified to produce calibrated streams of these trace metals into a simulated flue gas. The simulated flue gases are can be investigated in 3 unique reactor systems: (1) homogeneous flow, (2) packed-bed, and (3) entrained-flow, as shown in Figure 1. Adsorption isotherms can be determined by calculating the differences in the direct inlet and outlet trace metal concentrations of each reactor. An image of the flame in the burner is shown in Figure 2.
Figure 2. Image of the flame generated from methane oxidation for simulated flue gas experiments
Figure 2. Image of the flame generated from methane oxidation for simulated flue gas experiments
These tests are carried out over a series of fixed temperature conditions to obtain relevant adsorption isotherm data. The outlet flow is drawn into the electron ionization quadrupole mass spectrometer (EI-QMS) for the direct measurement of trace metal species. The flow reactor is used to investigate purely homogneous gas-phase trace metal chemistry, the packed-bed reactor is used to investigate adsorption and catalytic behavior of surfaces, and the entrained-flow reactor is used to take into account mass transfer effects by simulating entrained fly ash in the ductwork of a combustion utility. All reactors are made of quartz to minimize surface reactions and all stainless steel and Teflon tubing are Restek-treated to minimize potential interference reactions. The entrained-flow reactor consists of three different sections to conduct tests as a function of residence time by changing the reactor length. Each reactor section is covered with a ceramic fiber heater suitable for use up to 1100 °C. Sample particles are injected continuously at a constant rate through a syringe-type fluidized particle feeder. Experiments conducted in the Clean Air Conversion lab are capable of measuring trace metal speciation, adsorption and catalytic behavior of organic and inorganic materials in flue gas environment.

Funding Source: NSF (combustion)

Topic 2: Trace Metal Measurement and Speciation

The development of trace metal (TM) measurement techniques is important for understanding their speciation in flue gases and on the surfaces of flue gas solids. Of TMs, mercury (Hg) has been the most intensively studied in terms of its reactivity and continuous measurement; however, currently commercially available mercury analyzers are able to measure only elemental mercury, leaving the oxidized unspeciated. The amount of oxidized mercury is determined from the difference between the amount of elemental mercury and the amount of total mercury measured in a given flue gas stream. Sampling is typically performed using a sampling train, where the sample is passed through a series of aqueous solutions to separate and collect elemental mercury. The current mercury measurement techniques do not allow for distinguishing between the two different oxidized forms, Hg+ and Hg2+, which makes it difficult to understand mercury speciation. In addition, recent experimental studies reported that the mercury measurements performed with wet chemical conditioning systems can be biased due to side reactions involving the oxidized mercury (i.e., HgCl2 or HgBr2).

Given the shortcomings of the current Hg measurement techniques, it is essential to measure oxidized and elemental Hg directly to have a complete understanding of its speciation. An electron ionization quadrupole mass spectrometer (EI-QMS), which is specially designed for direct TM measurement, will be used to speciate the TMs in our experimental flue gas streams. The mass spectrometer consists of the ionizer, mass filter and detector. Gaseous molecules are ionized through the bombardment of electrons, and the ions are focused and accelerated in the quadrupole mass filter. Unstable ions are neutralized by the quadrupole mass filter, whereas stable ions are passed through and detected. Analysis is based upon the intensity of a given species as a function of its mass-to-charge ratio. A benefit of employing a mass spectrometer is, unlike commercially available Hg analyzers, the oxidized forms can be isolated and individually identified because it separates the products based upon their mass-to-charge ratio. Additionally, the EI-QMS can be used to directly speciate between all TM metal species so that arsenic and selenium and its compounds can be measured directly in addition to mercury. The Clean Energy Conversion lab is also equipped with a PS Analytical Sir Galahad mercury measurement device, which is used to benchmark and calibrate the mass spectrometer since it is the first of its kind to speciate trace metals in simulated combustion flue gases at ppb levels.

To accurately measure the low concentrations of different TM species present in coal combustion flue gases, the EI-QMS must be sensitive to concentrations in the ppb range. The EI-QMS's sensitivity depends largely on whether the molecular beam velocity is supersonic. To explain this dependence, the following section reviews the creation and nature of supersonic flow and its relationship to the EI-QMS's sensitivity. As the flow accelerates from a region of relative high pressure, P0, through an orifice into a region of lower background pressure, $P_b$, it will reach sonic speed if the exit pressure ratio $(P_0/P_b)$ exceeds a critical value $G$, defined by: $G=(\frac{y+1}{2})^y^/^(^y^-^1^)$ such that $y$, the heat capacity ratio, is defined as $f + \frac{2}{f}$ , where $f$ is the number of degrees of freedom of the molecule. If the pressure gradient is great enough to create supersonic free jet expansion, then the exit pressure of the flow becomes independent of $P_b$ and equals $P_0/G$ (thus exceeding $P_b$). The flow is considered underexpanded because it has a pressure higher than the background pressure of $P_b$; therefore, the flow expands to meet the necessary boundary conditions imposed by the background pressure. The core of this supersonic expansion, located in the 'zone of silence' region, is isentropic and unaware of any external conditions. Flow in the zone of silence is unaffected by the background gas because flow disturbances cannot propagate upstream faster than the supersonic speed of the flow. In regard to the EI-QMS design, the pressure gradient between the inlet chamber and the first vacuum chamber may be defined to create supersonic expansion. Then, with the skimmer located inside the zone of silence, the molecular beam will be extracted from the radially-confined isentropic flow. In such a setup, scattering of the molecular beam is avoided and the amount of gas that reaches the ionization region, and subsequently the ion detector, is maximized, thereby improving the sensitivity of the instrument.

Funding Sources: NSF (combustion); EPRI

Topic 3: Gas-Phase Reaction Kinetics

Gas-Phase Reaction KineticsSimilar to mercury (Hg), As and Se are volatile species released into the atmosphere from combustion processes. Their water-soluble forms have the potential to contaminate groundwater and particle-bound forms can deposit onto soil. Selenosis is a disorder known to affect livestock who graze on plants that uptake Se from the soil through their roots.

As environmental regulations pertaining to mercury become more stringent, other trace metals, particularly arsenic (As) and selenium (Se), are beginning to be examined as Hazardous Air Pollutants (HAPs) under the Maximum Achievable Control Technology (MACT) EPA standards. Similar to mercury (Hg), As and Se are volatile species released into the atmosphere from combustion processes. Their water-soluble forms have the potential to contaminate groundwater and particle-bound forms can deposit onto soil. Selenosis is a disorder known to affect livestock who graze on plants that uptake Se from the soil through their roots. Coal combustion flue gases and gasification fuel gases are known to contain compounds of these elements. As such, the mechanism for the removal of these pollutants during the combustion process is a topic of much attention. Any removal strategy will be dependent upon the speciation of these elements, and speciation will be dependent upon the reaction chemistry in the flue gas. Determination of the thermochemistry and kinetics of reactions involving these metals is essential for the development of effective removal techniques

Thermochemistry dictates which species are most stable and the reactions that are likely to occur. Experimental thermochemistry data is typically available only at standard temperature and pressure and in the case of some species, not at all. To accommodate for the limitations of experimental studies, theoretical computations can be conducted to calculate the thermochemical properties of different Hg, Se, and As species. Using first-principles quantum mechanics, the electronic structure of each species can be determined. From the electronic structure, optimized geometries, vibrational frequencies, enthalpies and entropies of reaction at all temperatures can be predicted. From enthalpy and entropy of reaction, Gibbs free energy of reaction and equilibrium constants can be calculated. Gibbs free energy and equilibrium constants indicate the thermodynamic stability of the compounds of interest.

Thermochemistry determines which species are most stable, but it cannot predict the timescales on which the reactions proceed. To comprehensively model trace metal speciation, it is necessary to calculate kinetic parameters for each reaction in addition to the thermochemistry. The derivation of rate expressions for the reactions in question can be used to determine the reactions that occur most rapidly. From this data concentration profiles of each trace metal species can be determined as a function of temperature. Within our group, rate expressions rate constants are calculated using conventional transition state theory (TST) for bimolecular reactions and RRKM theory for unimolecular reactions. Potential energy surfaces (PES) reveal how the potential energy varies as a function of the atomic or molecular coordinates of a given reaction. The PES has a minimum energy pathway, which corresponds to the reaction pathway, and a saddle point, which corresponds to the transition structure. From this information activation energies can be predicted and used alongside either TST or RRKM to determine rate constants. Kinetic parameters of the trace metal reactions are combined with other hydrocarbon reactions and incorporated into a global combustion model using CHEMKIN software. The global mechanism constitutes over 300 reactions including CH4, SOx, NOx, and halogen chemistry. Modeling results obtained from CHEMKIN are validated by comparison to the experimental results obtained in our combustion lab.

Funding Source: NSF (combustion)