Supplementary Material

Characterization of Trace Contaminants in Syngas from the

Thermochemical Conversion of Biomass

S. Kent Hoekman*1, Curtis Robbins1, Xiaoliang Wang1, Barbara Zielinska1,

Dennis Schuetzle2 and Robert Schuetzle3

1Division of Atmospheric Sciences, Desert Research Institute (DRI), 2215 Raggio Parkway, Reno, NV 89512

2Renewable Energy Institute International (REII), McClellan, CA 95652

3Pacific Renewable Fuels and Chemicals, McClellan, CA 95652

Part 1: Protocols for Sampling and Analysis of Syngas………………………………………2

Part 2: Results from Analysis of Syngas Samples Collected from Synterra IBR……….….15

Reference List...…………………………………………………………………………………22

Part 1: Protocols for Sampling and Analysis of Syngas

A. Canister Samples

1. Canister preparation and collection

Prior to use, electro-polished canisters are cleaned by alternating evacuation and flushing through seven cycles with humid, ultra high purity (UHP) air at 140ºC. Ten percent of the cleaned canisters are then pressurized with humid UHP air, allowed to equilibrate overnight, then analyzed by gas chromatography with flame ionization detection (GC/FID). For blank samples, each individual target compound should be present at concentration less than 0.2 ppbv; total non-methane hydrocarbon concentration should be less than 10 ppbC. Before shipping to a field site for use, the canisters are evacuated in the laboratory by connection to a vacuum line. When collecting syngas, the evacuated canister is opened by a solenoid valve, and the flow is regulated through a critical orifice.

2. Canister Analysis (for C2-C11)

For analysis of VOCs, a gas chromatography/mass spectrometry/flame ionization detector (GC/MS/FID) method is employed, based upon guidance provided by EPA Method TO-15.1 The integrated GC/MS/FID system includes a Lotus Consulting Ultra-Trace Toxics sample pre-concentration system built into a Varian 3800 GC instrument with a FID, coupled to a Varian Saturn 2000 ion trap mass spectrometer. The Lotus pre-concentration system consists of three traps. Mid- and heavier weight hydrocarbons are collected on the front trap consisting of 1/8” nickel tubing packed with multiple adsorbents. Trapping is performed at 55ºC and eluting is performed at 200ºC. The rear traps consist of an empty 0.040” ID nickel tubing (for light hydrocarbons) and a cryo-focusing trap (for mid and higher weight hydrocarbons isolated in the front trap). The cryo-focusing trap is built from 6’ x 1/8” nickel tubing filled with glass beads. Trapping in both rear traps occurs at -180ºC and eluting at 200ºC. Light hydrocarbons are deposited onto a Varian CP-Sil5 column (15m x 0.32mm x 1µm) plumbed to a column-switching valve within the GC oven, then to a Chrompack Al2O3/KCl column (25m x 0.53mm x 10µm) leading to the flame ionization detector (FID) for quantification of light hydrocarbons (C2-C4). The mid-range and heavier hydrocarbons cryo-focused in the rear trap are deposited onto a J&W DB-1 column (60m x 0.32mm x 1µm) connected to an ion trap mass spectrometer. The GC initial temperature is 5ºC, held for approximately 9.5 minutes, then ramped at 3ºC/min to 200ºC for a total run time of 80 minutes.

Calibration of the system is conducted with a mixture that contains commonly found hydrocarbons (75 compounds from ethane to n-undecane, purchased from Air Environmental) in the range of 0.2 to 10 ppbv. Three point external calibrations are run prior to analysis, and one calibration check is run every 24 hours. If the response of an individual compound is more than 10% off, the system is recalibrated.

Replicate analyses are conducted at least 24 hours after the initial analyses to allow re-equilibration of the compounds within the canisters. All replicate analyses are flagged in the project database. The data processing program extracts these replicates and determines a replicate precision. Replicate analysis is important because it provides a continuous check on all aspects of each analysis, and highlights analytical problems before they become significant.

Method detection limit (MDL) is determined as recommended by the EPA Method TO-15 (according to the Code of Federal Regulations, 40 CFR 136 Appendix B). Briefly, seven consecutive replicate measurements of the compounds of interest at concentrations near (within a factor of 5) the expected detection limits are made, and the standard deviations for these 7 replicate concentrations are calculated. The MDL is obtained by multiplying these standard deviations by 3.14 (i.e. the Student’s t-value for 99 percent confidence for 7 values). In general, MDLs for VOC measurements are 0.1 – 0.2 ppbv.

3. Canister Analysis of Permanent Gases (CO, CO2, N2, CH4, H2)

3a. Methane, Carbon Monoxide, and Carbon Dioxide

Methane (CH4), carbon monoxide (CO) and carbon dioxide (CO2) are measured from canister samples using GC/FID (Shimadzu GC-17A). Since the FID does not respond to CO and CO2, these species are converted to methane by a methanator, positioned immediately after the GC column, but ahead of the detector. The methanator comprises a firebrick powder impregnated with nickel catalyst, through which a stream of hydrogen gas flows continuously at ~450C.

For compound separation, a 20-ft x 1/8-in. inner-diameter (i.d.) column, packed with 60/80 mesh of Carboxen 1000 (Supelco) is used. This column provides sufficient separation between CH4 and CO without retaining CO2. Five-ml samples are injected using a constant volume loop. An initial column temperature of 35°C is held for 8-min., followed by a gradient of 15°C/min to a final temperature of 200°C. Response factors are determined by calibrations with gaseous standard mixtures (Scott Specialty Gases or AGA Specialty Gases, NIST-traceable) containing CO, CO2 and CH4 in zero air. The minimum detection limits are 0.06 ppmvfor CO,0.2 ppmv for CH4, and ~3 ppmvfor CO2 (MDLs determined as described above). The precision of measurements is generally better than 10%.

3b. Hydrogen and Nitrogen

Hydrogen (H2) and nitrogen (N2) concentrations are measured using a SRI 8610C gas chromatograph with a 0.5 ml sample loop and a thermal conductivity detector.A SRI Molecular Sieve 13x (6 ft x 1/8 in ID) GC column is used.The initial GC oven temperature is 40°C, followed by a temperature ramp of 50°C/min to 200°C and final hold of 6.8 minutes, giving a total run time of 10 minutes.

A single stock calibration standard was obtained from Airgas containing a blend of H2 (50.04% ±1%), CO (19.98% ±1%), CO2 (19.98% ±1%), and CH4 (10% ±1%).Lower calibration levels were prepared by quantitatively diluting the stock gas with ultra high purity N2 into 6-liter stainless steel canisters.H2 calibration levels were 7.48%, 20.28%, 30.17%, and 50.04%.N2 standards were prepared from an ultra high purity (UHP) N2 sample mixed with helium (He).

B. Carbonyl Samples

1. Carbonyl Collection

Carbonyl compounds are collected by drawing air through silica gel Sep-Pak® cartridges impregnated with acidified 2,4-dinitrophenylhydrazine (DNPH), available commercially from Waters, Inc. The resulting products (hydrazones) in the cartridges are measured in the laboratory using high performance liquid chromatography (HPLC). The dilution sampler system used for carbonyl collection includes check valves, solenoid valves and a pump to enable flow control and measurement for each cartridge. Used cartridges are immediately removed from the sampler, plugged, put into vials, and stored in a refrigerator until they are returned to the laboratory (in a cooler) for final analysis.

2. Carbonyl Analysis

After sampling, the DNPH Sep-Pak® cartridges are eluted with 2-mL acetonitrile to remove the hydrazone products produced during sampling of carbonyl compounds. An aliquot of the eluent is transferred into a 2-mL septum vial and injected with an autosampler into a high performance liquid chromatograph (HPLC; Waters 2690 Alliance System with 996 Photodiode Array Detector) for separation and quantification of the hydrazones.2The chromatographic conditions are as follows: Polaris C18-A 3µm 100 x 2.0 mm HPLC column, flow rate of 0.2 ml/min, injection volume of 2.0 µl, solvent A: water, solvent B: acetonitrile. The HPLC program is: 60% A, 40% B for 0.02 min., 50% A and 50% B over 15 min., 30% A and 70% B over 6 min., and 100% B over 1 min., final hold at 100% B for 1 min. Total run time is 30 min. C1 through C7 carbonyl compounds are analyzed, including the following: formaldehyde, acetaldehyde, acetone, acrolein, propionaldehyde, crotonaldehyde, methyl ethyl ketone, methacrolein, butyraldehyde, benzaldehyde, glyoxal, valeraldehyde, m-tolualdehyde, and hexanaldehyde. The original carbonyl concentrations in the syngas (in units of ppbv) are computed from the amounts measured after blank correction, and after accounting for the volume of syngas sampled. MDLs are determined according to the method described above for VOC, and are generally in the range of 0.1-0.2 ppbv.

C. Tenax C8-C20VOC Samples

1. Tenax Collection

Tenax sampling and analysis is employed for compounds that are too heavy to be quantitatively retrieved from canisters. [These higher MW VOCs are sometimes called semi-volatile organic compounds (SVOCs).] Prior to use, the Tenax-TA solid adsorbent is cleaned by a Dionex Accelerated Solvent Extractor(ASE) method using hexane/acetone mixture (4/1 v/v), and dried in a vacuum oven at ~80ºC. The dry Tenax is packed into Pyrex glass tubes (4 mm i.d. x 15 cm long; each tube containing 0.2 g of Tenax) and is thermally conditioned for four hours by heating in an oven at 300ºC under a nitrogen purge (25 mL/min N2 flow). Approximately 10% of the pre-cleaned Tenax cartridges are tested by GC/MS for quality assurance prior to sampling. After cleaning, the Tenax cartridges are capped tightly using clean Swagelok caps (brass) with graphite/vespel ferrules, placed in metal containers with activated charcoal on the bottom, and kept in a clean environment at room temperature.

As described above for the DNPH cartridges, flow control and measurement is done for each individual Tenax cartridge during use. After use, the exposed cartridges are removed, immediately plugged with Swagelok caps, and stored in the same metal containers with activated charcoal on the bottom. The exposed cartridges are stored under refrigeration until they are returned to the laboratory in a cooler containing blue ice.

2. Tenax Analysis

The Tenax samples are analyzed by a thermal desorption-cryogenic pre-concentration method, followed by high-resolution gas chromatographic separation and mass spectrometric detection (GC/MS) of individual compounds.3A GerstelThermoDesorption System (TDS) unit, equipped with a 20 position autosampler is used for these analyses. This TDS unit is coupled to a Varian Saturn 2000 GC/MS system, equipped with sample desorption and cryogenic pre-concentration capabilities. A DB-1 capillary column (60 m x 0.32 mm i.d., 0.25 µm film thickness; J&W Scientific, Inc.) is used to achieve separation of the target species. The GC initial temperature of 30°C is held for 3-min., then increased to 250°C at 5°C/min., and held for 3-min. for a total run time of 50 min.

For calibration purposes, Tenax cartridges are spiked with a methanol solution of standard hydrocarbons, prepared from high-purity commercially available C8-C20 aliphatic, oxygenated, and aromatic hydrocarbons. The solvent is then removed with a stream of He (2 min, 100 mL/min at room temperature) and the Tenax cartridges are thermally desorbed into the GC system. Three concentrations of each standard compound are employed and two replicate sample injections per calibration level are made. Area response factors per nanogram of compound per Tenax cartridge are calculated by the instrument software for each concentration. All response factors are recorded in the software program and the mean or median value is taken. The original concentrations of SVOCs in the syngas (expressed in units of µg/m3) are computed after accounting for the volume of syngas sampled. MDLs are determined as described above for VOC, and are generally in the range of 0.01-0.02 µg/cartridge.

D. Filter Packs

1. Filter Sample Collection

No single filter medium is appropriate for all desired analyses, so it is necessary to sample on multiple substrates for chemical speciation. Filter packs containing 47-mm diameter Teflon-membrane, quartz fiber, and cellulose filters are used for syngas sampling and analysis. All filter batches are conditioned and acceptance tested prior to use in sampling. Two percent of filters from each batch are subjected to identical blank analyses, to ensure that they are sufficiently clean before use in actual sampling. The following three types of filters are used:

  1. Teflon-membrane filters are used for measurement of mass and elemental concentrations. These filters are obtained from Pall Corporation (Part No. R2PJ047) or Whatman Inc. (Part No. 7592-104). They have a 2 µm pore size, and are used with polymethylpropylene (PMP) support rings.
  2. Quartz fiber filters are used for determination of carbon fractions and ions in the particulate phase. These filters are obtained from Pall Corporation (Part No. 7202) or Whatman Inc. (Part No. 1851-047). They have an approximate thickness of 432 µm, fiber diameter of 0.6 µm, and packing density of 0.038 g/cm3.
  3. Cellulose fiber filters are placed behind the more efficient particle-collecting filters (Teflon-membrane and quartz fiber). They are impregnated with gas-absorbing compounds, and are used to capture ammonia (with citric acid impregnation), acidic gases (with K2CO3 impregnation), and H2S (with AgNO3 impregnation). These filters are obtained from Whatman Inc. (31ET and 41), and have a thickness of 0.50 mm.

2. PM Mass by Gravimetric Analysis

Unexposed and exposed Teflon-membrane filters are equilibrated at a temperature of 21.5 ± 1.5 °C and a relative humidity of 35 ± 5% for a minimum of 24 hours prior to weighing. Weighing is performed on a Metter Toledo MT5 electro-microbalance with ±0.001 mg sensitivity. The charge on each filter is neutralized by exposure to a 210Po ionizingsource for 30 seconds or more prior to the filter being placed on the balance pan. The balance is calibrated with a series of three Class 1 weights (50, 100, and 200 mg) and the tare is set prior to weighing each batch of filters. After every 10 filters are weighed, the calibration and tare are re-checked. If the results of these performance tests deviate from specifications by more than ±5g, the balance is re-calibrated.

Replicate weights are determined on 100% of the filters before sampling (initial weights or pre-weights), and on 30% of the filters after sampling (final weights or post-weights) by an independent technician. Replicate pre-sampling (initial) weights must be within ± 0.010 mg of the original weights. Replicate post-sampling (final) weights on lightly-loaded samples (i.e., less than 1 mg) must be within ± 0.015 mg. Post-sampling weights on heavily loaded (i.e., greater than 1 mg) samples must be within 2% of the net weight. Pre- and post-weights, check weights, and re-weights (if required) are recorded on data sheets as well as being directly entered into a database via an internet connection.

3. Elements by X-Ray Fluorescence

Individual elements are analyzed on Teflonmembrane filters using a PANalytical Epsilon 5, energy dispersive x-ray fluorescence (ED-XRF) analyzer.4 The emissions of x-ray photons from the sample are integrated over time and yield quantitative measurements for 51 elements ranging from aluminum (Al) through uranium (U), and semi-quantitative measurements of sodium (Na) and magnesium (Mg). A spectrum of x-ray counts versus photon energy is acquired and displayed during analysis, with individual peak energies corresponding to each element, and peak areas corresponding to elemental concentrations. The advantages of XRF analysis include high sensitivity for a large number of elements, the ability to analyze small sample quantities, and the non-destructive nature of the analysis.

The source of x-rays in the PANalytical Epsilon 5 analyzer is a sidewindow, liquidcooled, 100 KeV, 24 milliamp gadolinium anode xray tube. X-rays are focused on one of 11 secondary targets (Al, Ca, Ti, Fe, Ge, Zr, Mo, Ag, Cs, Ba, Ce) which in turn emit polarized x-rays to excite a sample. Xrays from the secondary target or the tube are absorbed by the sample, exciting electrons to higher level orbitals. As the electrons return to their ground state, photons are emitted which are characteristic of the quantum level jumps made by the electron; the energy of the emitted photons are, therefore, characteristic of the elements contained in the sample. The fluoresced photons are detected in a solid state germanium x-ray detector. Each photon that enters the detector generates an electrical charge whose magnitude is proportional to the photon's energy. The number of these photons is proportional to the number of atoms present. Ten separate XRF analyses are conducted on each sample to optimize detection limits for the specified elements. The ED-XRF system is calibrated using Micromatter (Arlington, WA) thin film standards. Multielement standards are analyzed daily to monitor for any instrument drift. Method detection limit (MDL) is defined as 3 times the standard deviation of multiple measurements of a laboratory blank filter.

4. Carbon Analysis by Thermal/Optical Reflectance/Transmittance(TOR/TOT)

The thermal/optical reflectance and transmittance (TOR/TOT) method measures organic carbon (OC) and elemental carbon (EC) on filter samples.5 This method is based on the principle that different types of carbon-containing particles are converted to gases under different temperature and oxidation conditions. The different carbon fractions from TOR/TOT are useful for comparison with other methods, which are specific to a single definition for organic and elemental carbon. The seven carbon fractions measured by the DRI Model 2001 Thermal/Optical Carbon Analyzer are the following:

1)OC1: Carbon evolved in a helium atmosphere at temperatures between ambient and 140 °C

2) OC2: Carbon evolved in a helium atmosphere at temperatures between 140 and 280 °C

3)OC3: Carbon evolved in a helium atmosphere at temperatures between 280 and 480 °C

4)OC4: Carbon evolved in a helium atmosphere between 480 and 580 °C

5)EC1: Carbon evolved in an oxidizing atmosphere at 580 °C

6)EC2: Carbon evolved in an oxidizing atmosphere between 580 and 740 °C

7)EC3: Carbon evolved in an oxidizing atmosphere between 740 and 840 °C

The thermal/optical reflectance carbon analyzer contains a thermal system and an optical system. The thermal system consists of a quartz tube placed inside a coiled heater. Current through the heater is controlled to attain and maintain pre-set temperatures for given time periods. A portion of a quartz filter is placed in the heating zone and heated to different temperatures under non-oxidizing and oxidizing atmospheres. The optical system consists of a He-Ne laser, a fiber optic transmitter and receiver, and a photocell. The filter deposit faces a quartz light tube so that the intensity of the reflected laser beam can be monitored throughout the analysis.

As the temperature increases from ambient (~25 °C) to 580 °C, organic compounds are volatilized from the filter in a non-oxidizing (He) atmosphere while elemental carbon is not oxidized. When oxygen is added to the helium at temperatures greater than 580 °C, the elemental carbon burns and enters the sample stream. The evolved gases pass through an oxidizing bed of heated manganese dioxide where they are oxidized to carbon dioxide, then across a heated nickel catalyst, which reduces the carbon dioxide (by reaction with hydrogen) to produce methane (CH4). The methane is then quantified with a flame ionization detector (FID).