Qualitative Characterization of the Aqueous Fraction from Hydrothermal Liquefaction of Algae Using 2D Gas Chromatography with Time-of-flight Mass Spectrometry

Summary

A two-dimensional gas chromatography-time-of-flight mass spectrometry method is described for characterization of the aqueous fraction of bio-crude produced from hydrothermal liquefaction of algae. This protocol can also be employed to analyze the aqueous fraction of liquid products from fast pyrolysis, catalytic fast pyrolysis, catalytic deoxygenation and hydro-treating.

Abstract

Two-dimensional gas chromatography coupled with time-of-flight mass spectrometry is a powerful tool for identifying and quantifying chemical components in complex mixtures. It is often used to analyze gasoline, jet fuel, diesel, bio-diesel and the organic fraction of bio-crude/bio-oil. In most of those analyses, the first dimension of separation is non-polar, followed by a polar separation. The aqueous fractions of bio-crude and other aqueous samples from biofuels production have been examined with similar column combinations. However, sample preparation techniques such as derivatization, solvent extraction, and solid-phase extraction were necessary prior to analysis. In this study, aqueous fractions obtained from the hydrothermal liquefaction of algae were characterized by two-dimensional gas chromatography coupled with time-of-flight mass spectrometry without prior sample preparation techniques using a polar separation in the first dimension followed by a non-polar separation in the second. Two-dimensional plots from this analysis were compared with those obtained from the more traditional column configuration. Results from qualitative characterization of the aqueous fractions of algal bio-crude are discussed in detail. The advantages of using a polar separation followed by a non-polar separation for characterization of organics in aqueous samples by two-dimensional gas chromatography coupled with time-of-flight mass spectrometry are highlighted.

Introduction

Steady growth in demand for liquid fuels, finite fossil fuel resources, uncertainty of fossil fuel supplies, and concerns over the increasing concentration of greenhouse gases in the atmosphere have increased global awareness for renewable resources¹. Solar energy (including photovoltaics and solar-thermal), wind energy, hydropower, geothermal, and biomass are the primary renewable sources that could potentially replace fossil-derived energy². Of these, biomass is the only carbon-based alternative energy resource for the production of liquid transportation fuels and high-value chemicals³. Biomass includes any organic material such as forest resources, agricultural residue, algae, oilseeds, municipal solid waste, and carbon-rich industrial wastes (e.g. from pulp and paper industry or from food processing)¹. Biomass is classified into two broad categories: lignocellulosic and non-ligneous feedstocks based on compositional characteristics. Lignocellulosic biomass consists of carbohydrates and lignin, while non-ligneous feedstocks have proteins, carbohydrates and lipids/oils⁴. Lignocellulosic feedstocks, derived from terrestrial plants, can only satisfy 30% of the current liquid fuel (gasoline, jet fuel, and diesel) demand if sustainably cultivated and harvested^5,6. Hence, non-ligneous aquatic microorganisms, such as microalgae and fungi, are considered potential feedstocks for the production of renewable liquid fuels to complement lignocellulosic resources.

Microalgae feedstocks have the potential to satisfy current liquid transportation fuels demand^7,8. Algae have many advantages: high areal productivity⁸, the ability to grow in low-quality, brackish, or sea water⁹, and the ability to accumulate energy-dense triglycerides or hydrocarbons^7,8. Hydrothermal liquefaction (HTL) is a viable and scalable conversion process which utilizes water naturally associated with algal or aquatic feedstocks^10,11. It is a thermo-chemical process with operating temperatures of 250-400 °C and operating pressures of 10-25 MPa which produces a liquid product, or bio-crude, which can be upgraded into a fuel blend stock. Bio-crude produced from HTL of algae has distinguishable and easily separable organic and aqueous fractions. The organic fraction of bio-crude can be efficiently converted into a refinery ready blend stock via catalytic hydro-treating processes¹¹. The aqueous fraction of bio-crude contains ~30% of the total carbon present in the algal feedstock. Although thousands of compounds have been identified in the HTL aqueous stream, the predominant fractions consist of low molecular weight oxygenates (including acids, alcohols, ketones, and aldehydes) formed by the degradation of carbohydrates and lipids, and nitrogen heterocyclics (including pyrroles, pyridines, pyrazines, and imidazoles) derived from protein decomposition¹². Studies on utilizing the aqueous fraction to improve overall process economics as well as sustainability are ongoing. Synthesis gas can be produced from the aqueous fraction of algae bio-crude via catalytic hydrothermal gasification^10,13,14. Alternatively, organics in the aqueous fraction can also be catalytically converted to fuel additives and specialty chemicals. Research on optimizing catalytic hydrothermal gasification and catalyst screening studies for conversion of organics in the aqueous liquid phase is currently underway at the Pacific Northwest National Laboratory (PNNL). For this work, qualitative as well as quantitative characterization of the aqueous fraction of algae bio-crude is required. Since the aqueous fraction of algae bio-crude is considered a waste stream, there are very few studies that have analyzed the aqueous fraction of algae bio-crude^13,15. Moreover, recent studies concluded that converting this HTL algae water into high-value bio-products would improve the sustainability as well as economics of an HTL-based bio-refinery¹¹. Therefore, this study focused on developing a method for qualitative characterization of the aqueous fraction of bio-crude obtained from HTL of algae by two-dimensional gas chromatography coupled with time-of-flight mass spectrometry (GC × GC–TOF-MS).

GC × GC–TOF-MS is the most promising chromatographic analytical technique to increase resolution (or separation of chemical compounds in a sample), peak capacity (i.e. number of resolved peaks), signal-to-noise ratio (for identification of chemical compounds with high confidence), and to avoid co-elution of chemical compounds¹⁶. In order to maximize resolution, peak capacity, and signal-to noise ratio, two GC columns with different stationary phases are connected in series using a press-fit connector or micro-union¹⁷(see Figure 1 which is a block diagram of GC × GC–TOF-MS system used in this study). A modulator is located between the press-fit connector and secondary columns to trap, refocus, and re-inject the effluents from the primary column into the secondary column¹⁸. Modulation occurs on the secondary column in the present study as shown in Figure 1. The secondary column is then connected to the TOF-MS via a transfer line assembly.

GC × GC–TOF-MS was used previously for qualitative as well as quantitative analysis of organic samples such as crude oil^16,19, gasoline, jet-fuel, diesel, bio-diesel, and the organic fraction of bio-fuel^20-22 produced from thermo-chemical as well as thermo-catalytic conversion processes^23,24. For characterization of these organic samples in GC × GC–TOF-MS instruments, a long non-polar column was used as the primary column, while a short polar column was used as the secondary column. This conventional column configuration resolves chemical compounds based on differences in volatility over the first dimension, followed by polarity in the second dimension¹⁸. Aqueous or water samples from biological processes, food processing, and environmental wastes were also characterized using similar primary/secondary column configurations after the sample had been through preparation steps^17,25-30. Sample preparation techniques such as derivatization, solid-phase extraction, and organic solvent extraction have all been utilized prior to GC × GC–TOF-MS analysis^{17,27-29,31,32}. These techniques were aimed at decreasing the polarity of compounds in the sample for analysis using a conventional column configuration³³. An alternative strategy was employed in this study based on the nature of the sample (here polar organic compounds in water) utilizing the reverse primary/secondary column configuration for GC × GC–TOF-MS analysis. Since the aqueous fraction of bio-crude produced from HTL has polar compounds¹³, a column combination of a primary polar column and a secondary non-polar column was used in the GC × GC–TOF-MS without any upstream sample preparation. This primary/secondary column combination resolves chemical compounds based on differences in polarity over the first dimension, followed by volatility in the second dimension. Limited analytical methods exist in the literature for characterization of aqueous samples using two-dimensional gas chromatography without prior sample processing¹⁵.

The objective of this study was to determine the chemical compounds present in the aqueous fraction of algae bio-crude. To achieve this objective, a GC × GC–TOF-MS data acquisition method was developed with a column combination of polar column (primary) × non-polar (secondary). Klenn et al. (2015) suggested that increasing the length of the primary column (especially 60 m GC columns) and lowering the offset temperature of the secondary column with respect to the primary column would maximize peak capacity and resolution^16-18. Therefore, a 60 m primary column and 5 °C offset temperature of the secondary column with respect to the primary column were used in this study. The optimum modulation period was determined following a protocol described in this study (see section 4). The optimum ramp rate of GC column temperature was determined by a trial and error method and is similar to the value suggested in the literature^16-18. To discuss the advantages of this column combination for aqueous samples, we have analyzed HTL algae water samples with the conventional column combination of non-polar × polar. Operating parameters suggested in the literature were employed for analyzing the aqueous fraction of algal bio-crude with a non-polar × polar column combination¹⁸.

Protocol

1. Sample Preparation

1. Generate a mixed aqueous/organic product stream via continuous flow HTL of algae according to the reactor design and experimental procedure found in the literature^10,11.
2. Use a gravity separator to separate the product stream into an aqueous phase and organic phase.
3. Filter 10 ml of the HTL aqueous phase using a 0.45 µm syringe filter and store in a refrigerator maintained at 4 °C for GC × GC–TOF-MS analysis.

2. Instrument Components

1. Use a gas chromatograph (GC) equipped with a quad-jet dual stage cooling-based modulator and time-of-flight (TOF) mass spectrometer (MS) for these experiments.
2. Configure the auto-sampler to inject 1 µl of each sample or standard into the GC. Use a randomized block design of sample and standard injections for the auto-sampler sequence as described in the literature¹³. The randomized block design is commonly used in quantitative studies to control for instrument operation. Our laboratory utilizes the design routinely even in comparative studies to verify instrument operation.
3. Connect the primary and secondary column using a press-tight connector before the modulator. Ensure that both edges of both primary and secondary columns are cut straight without sharp edges before connecting to the press-tight connector.
4. Place ferrule on the GC column and then connect primary column to the GC injector so that 5 mm of column is inside the injector.
5. Ensure that glass liner, non-stick liner O-ring and septa for GC injector are new and free of contamination.
6. Use 1/16 x 0.5 mm ID transfer line ferrules to connect the secondary column and transfer line. Place a 0.2 m portion of the secondary column in the transfer line.
7. Ensure that a 0.1 m portion of the secondary column is in the modulator.
8. Use ultrahigh purity helium gas as carrier gas for GC at a flow rate of 1.5 ml min^-1.
9. Ensure there is sufficient liquid nitrogen in the Dewar which acts as a coolant in the modulator. The level of the liquid nitrogen in the Dewar can be predicted using a pressure gauge attached to its outlet. A 69 kPa reading of the pressure gauge indicates that the Dewar is full, while 0 kPa indicates that it is empty.

3. Protocols Before Analyzing Samples

1. Ensure there are no major leaks in the instrument. If the vacuum gauge reading of the TOF-MS is higher than 2.7 × 10^-5 Pa for 1.5 ml min^-1 GC column flow rate, this indicates a major leak in the system.
2. Set-up the quality control (QC) method and run in-built 'acquisition system adjustments' protocol to achieve maximum signal response using manufacturer's protocol.
3. Run in-built 'instrument optimization' protocols of QC method, in series - filament focus, ion optic focus and mass calibration tests using manufacturer's protocol. Ensure that mass calibration test passes. This QC method ensures that all the hardware parameters of the instrument are at optimum level.
4. Perform a "leak check" using manufacturer's protocol. Analyze automatically generates leak check report. Ensure that the relative concentration of 28 (N₂), 32 (O₂) and 18 (moisture) ions must be below less than 10%, 3% and 5% of internal standard mass spectra of 69 ion, respectively.
5. Tune the TOF-MS using manufacturer's protocol.
6. Run quality control method as well as TOF-MS tune protocol before and after leak check and also while analyzing samples and standards.

4. Protocol to Determine the Optimum Modulation Period of Modulator

1. Arbitrarily select a long modulation period (e.g. 10 sec or 13 sec). Inject a sample as described in 2.2.
2. Identify the retention time in second dimension of the contour plot after which no peaks elutes. Select identified second dimension retention time as optimum modulation period. Figure 2 clearly elucidate the identification of retention time in second dimension of the contour plot.
3. Increase the modulation period used in step 4.1 and perform the analysis again if "wrap around" is observed¹⁸. Wrap around phenomena occurs if the peaks in the second dimension elutes below the baseline of first dimension. Example contour plot for 'wraparound' is shown in supplementary information Figure 3.
4. Repeat steps 4.2 and 4.3 until optimum value is determined.

5. Experimental Parameters of Instrument Set-up

1. Install a polar (60 m x 0.25 mm x 0.5 µm film thickness) capillary column as the primary column and a non-polar (2.3 m x 0.25 mm x 0.5 µm film thickness) capillary column as the secondary column. Bake both the primary and secondary column for at least 2 hr to remove trace amounts of moisture, air and contaminants associated with new GC columns.
2. Use ultrahigh purity helium gas as carrier gas for GC at a flow rate of 1.5 ml min^-1.
3. Set the GC injector to a temperature of 260 °C and a split ratio of 1:250.
4. Employ the following temperature program for the primary column: a constant temperature of 40 °C for 0.2 min followed by a temperature ramp to 260 °C at 5 °C min^-1, followed by a constant temperature of 260 °C for 5 min.
5. Maintain the modulator temperature 5 °C higher than that of the secondary column and the secondary column temperature at 5 °C higher than that of the primary column.
6. Use an optimum modulation period of 4 sec with 0.8 sec of hot pulse and 1.2 sec of cold pulse. This value is determined based on the protocol described in section 4.
7. Set transfer line temperature to 270 °C.
8. Set the acquisition delay or solvent delay to 0 sec.
9. Set the lower and higher range of m/z as 35 and 800, respectively.
10. Set the MS detector acquisition rate at 400 spectra/sec.
11. Maintain the MS detector voltage at 150 V higher than the optimized value.
12. Maintain the MS ion source temperature at 225 °C.

6. Data Analysis

1. Perform data processing using the software supplied by the instrument manufacturer.
2. Select the following tasks in the data analysis method - Compute baseline, find peaks above the baseline, library search and calculate are/height.
3. Track the baseline through the data file. Enter baseline offset as 0.5.
   Enter expected peak width of 15 sec in the first dimension and 0.15 sec in the second dimension.
4. Set signal-to-noise ratio as 5,000 and similarity values of >850 for identification of compounds.
5. Select a commercially available mass spectral library to identify chemical compounds present in samples and set the library search mode to forward.
6. Process the data files using this data analysis method using manufacturer's protocol. It requires at least 1 hr to process a data file.

Results

A total ion chromatogram (TIC) obtained for the aqueous fraction of algae bio-crude analyzed with a column combination of polar × non-polar is shown in Figure 4. Retention times and similarity or match factor values of compounds identified by searching against a National Institute of Standards and Technology (NIST) library are tabulated in Table 1. Oxygenates (such as cyclopenatanone, furanic compounds and dianhydromannitol) and organic acids (includ...

Discussion

Results clearly illustrate the ability of the column combination of polar × non-polar to resolve polar compounds and light volatiles present in the aqueous fraction of algae bio-crude without prior sample preparation techniques. Drastic peak tailing was observed for organic acids and N-compounds while using the non-polar× polar column combination. This peak tailing was not observed for the early eluting light organics. This behavior has been reproducible when verifying the instrument is free of leaks (the vacuu...

Materials

Name	Company	Catalog Number	Comments
GC × GC–TOF/MS	Leco	PEG4D11DLN15	Commercial Pegasus 4D
ChromaTOF version 4.50	Leco		Data analysis software
Rxi-5MS GC column	Restek	13420	2.3 m column was used from this column.
Stabilwax GC column	Restek	10626
HP-5 GC column	Agilent	19091J-416
Stabilwax GC column	Restek	15121
Presstight Connector	Restek	20430
GC injector liner	Restek	23305.5
GC Injector ferrules	Agilent	5181-3323
Non-stick liner O-rings	Agilent	5188-5365
Transfer line ferrules	Restek	20212
Ethanol	Sigma-Aldrich	459844	Chromatography grade
Acetone	Sigma-Aldrich	414689	Chromatography grade
Acetic acid	Sigma-Aldrich	320099	Chromatography grade
2-butanone	Sigma-Aldrich	360473	Chromatography grade
Propanoic acid	Sigma-Aldrich	402907	Chromatography grade
Butanoic acid	Sigma-Aldrich	19215	Chromatography grade
Pyridine	Sigma-Aldrich	270970	Chromatography grade
Pyrazine	Sigma-Aldrich	65693	Chromatography grade
Acetamide	Sigma-Aldrich	695122	Chromatography grade
2,5-pyrrolididione	Sigma-Aldrich	S9381	Chromatography grade
N-methylsuccinimide	Sigma-Aldrich	325384	Chromatography grade
N-(2-hydroxyethyl)succinimide	Sigma-Aldrich	444073	Chromatography grade