Molecular Characterization of Inhibiting Biochar Water-ExtractableSubstances Using Electrospray Ionization Fourier Transform IonCyclotron Resonance Mass SpectrometryCameron R. Smith, Rachel L. Sleighter, Patrick G. Hatcher, and James W. Lee*

Department of Chemistry and Biochemistry, Old Dominion University, Norfolk, Virginia 23529, United States

*S Supporting Information

ABSTRACT: Biochar has gained significant interest worldwide for its potentialuse as both a carbon sequestration technique and soil amendment. Recently,research has shown that pinewood-derived biochar water extracts inhibited thegrowth of aquatic photosynthetic microorganisms, both prokaryotic andeukaryotic algae, while chicken litter- and peanut shell-derived biochar waterextracts showed no growth inhibition. With the use of electrodialysis, thepinewood-derived biochar water extract is separated into 3 fractions (anode-isolated, center chamber retained, and cathode-isolated substances) all withvarying toxic effects. Because of its ultrahigh resolution and mass precision,electrospray ionization (ESI) coupled to Fourier transform ion cyclotronresonance mass spectrometry (FTICR-MS) is utilized in this study to analyzebiochar water extracts at a molecular level to enhance our understanding of thetoxic nature of pinewood-derived biochar water extracts as compared to benignpeanut shell-derived biochar water extracts. The molecular composition ofpinewood-derived biochar water extracts shows unique carbohydrate ligneous components and sulfur containing condensedligneous components that are both absent from the peanut shell water extracts and more prevalent in the anode-isolatedsubstances. Using Kendrick mass defect analysis, we also determine that the most likely inhibitor species contain carboxyl andhydroxyl homologous series, both of which are characteristic functional groups hypothesized in our previous research for theinhibitor species. We have suggested that inhibition of aquatic photosynthetic microorganism growth is most likely due todegraded lignin-like species rich in oxygen containing functionalities. From the study conducted here, we show the potential ofultrahigh resolution FTICR-MS as a valuable analytical technique for determining whether certain biochars are safe and benignfor use as carbon sequestration and soil amendment.

■ INTRODUCTIONThe world currently faces a systemic energy and environmentalproblem of increasing CO2 emissions and global climatechange. Solving this problem requires a comprehensiveportfolio of research and development efforts with multiplestrategies. Among these, the approach of using advancedbiomass pyrolysis with reduced emissions to produce biocharsoil amendments, in addition to biosyngas and biofuels, appearsto be a particularly promising pathway to sustainability. Thecentral idea is that biochar, produced cleanly and sustainably bypyrolysis of biomass wastes and used as a soil amendment,would “lock up” biomass carbon in a form that can persist insoils for hundreds to thousands of years; while at the sametime, help to retain nutrients in soils and reduce the runoff ofagricultural chemicals. This approach is receiving increasedworldwide attention.1,2 Globally, each year about 6.6 gigatons(Gt) of dry matter biomass3 (e.g., crop stovers, dead leaves,waste woods, and rice straws) are produced but not effectivelyutilized. Application of this approach could transform thesetypes of waste into valuable biochar, biosyngas, and biofuelproducts at Gt scales in a distributed manner. Worldwide, thisapproach could result in a net reduction of greenhouse gas

emissions by about 1.8 Gt of CO2−C (GtC) equivalentemissions per year, which is about 12% of the current globalanthropogenic emissions.4 This is a unique “carbon-negative”bioenergy system approach, which on a life-cycle basis couldnot only reduce but also reverse the human effects on climatechange.5 However, many more studies are necessary before thisapproach can be considered for widespread implementation.First, biochar occasionally shows inhibitory effects on plantgrowth.6,7 Organic species, including possibly inhibitory andbenign (or stimulatory) chemicals,8,9 are produced as part ofthe biomass pyrolysis process.10,11 Thus far, very little is knownabout the mechanisms of these factors. If biochar were to beglobally utilized as a soil amendment and carbon sequestrationagent at Gt C scales, the release of potentially toxic compoundsinto soils and associated hydrological systems may havenegative consequences, in both agro-ecosystems and aquaticenvironmental systems. It is essential to understand and

Received: August 5, 2013Revised: October 25, 2013Accepted: November 1, 2013Published: November 1, 2013

Article

pubs.acs.org/est

© 2013 American Chemical Society 13294 dx.doi.org/10.1021/es4034777 | Environ. Sci. Technol. 2013, 47, 13294−13302

mitigate any undesirable effects in order for biochar to beutilized as a soil amendment and carbon sequestration agent atGtC scales.Recently, we conducted a bioassay study12 showing that

some biochar water-extractable substances (BWES) maycontain certain chemical species toxic to aquatic photosyntheticmicroorganisms, including both blue-green alga (cyanobacteriaSynechococcus) and eukaryotic green alga (Desmodesmus). Giventhe potential environmental importance of biochar as a carbonsequestration technique and its possible BWES toxicity, it isimportant that the characterization of the BWES be thoroughlyinvestigated before they are introduced into the environment.According to a recent study done by Lin et al.13 using liquidchromatography−organic carbon detection (LC−OCD), bio-char water-extracts can be extremely complex containing lowmolecular weight acids, biopolymers, humic substances, andother components. Fourier transform ion cyclotron resonancemass spectrometry (FTICR-MS) to characterize biochar hasrecently become of interest to several research groups.Podgorski et al.14 coupled desorption atmospheric pressurephotoionization (DAPPI) with FTICR-MS for the directcharacterization of natural organic matter (NOM) for bothcombusted and pyrolyzed biochar derived from oak. Analysisshowed that combusted oak compounds with high O/C andH/C ratios were attributed to incomplete thermal degradationof lignin and hemicellulose, while pyrolyzed oak showed asignificantly larger amount of compounds with lower O/C andH/C ratios, characteristic of aromatic compounds. Mostrecently, Cole et al.15 studied the biochar entrapped organiccompounds during pyrolysis and gasification of switchgrass-derived biochar using laser desorption ionization (LDI) andatmospheric pressure photoionization (APPI) FTICR-MS forhydrophobic compounds, and electrospray ionization (ESI)FTICR-MS for hydrophilic compounds. Similar to Podgorski etal.,14 Cole et al.15 found that pyrolyzed biochar extractscontained phenolic- and carbohydrate-derived compoundsattributed to the pyrolysis of lignin and holocellulose, whilethe gasification produced a significantly larger amount ofpolycyclic aromatic hydrocarbons (PAHs).Here, we report the chemical analysis of BWES with ESI-

FTICR-MS to gain insight into the molecular levelcomposition. Because of the ultrahigh resolving power andmass accuracy of FTICR-MS at high magnetic fields, thecharacterization of two different BWES exhibiting oppositeeffects on algal growth, pinewood BWES toxic effect andpeanut shell BWES benign effect, can be compared. Ourprevious study12 showed evidence that the hypothesized toxiccompound(s) most likely are small (less than 500 Da) phenolicspecies with at least one carboxyl group. Both carboxyl andhydroxyl groups ionize efficiently in negative ion mode ESI, andthus, ESI-FTICR-MS is ideal for analysis of these BWESspecies. The objectives of this study are to investigatepinewood-derived BWES and their toxicity on the growth ofaquatic microorganisms at a molecular level. Molecularcharacterization of the water-soluble components associatedwith biochar is extremely important because of the likelihood ofthese species entering aquatic ecosystems in a variety of ways(i.e., leaching and runoff), if used as a soil amendment andcarbon sequestration agent at GtC scales. This study will alsodemonstrate the effectiveness of FTICR-MS analysis and itsapplication toward the production of environmentally friendlybiochar.

■ MATERIALS AND METHODS

Biochar Material. Three different biochar samples wereutilized in this study: (1) peanut shell-derived biochar, (2)chicken litter-derived biochar, and (3) pinewood-derivedbiochar. The peanut shell and chicken litter-derived biocharmaterials were produced from pelletized peanut shells andchicken litter through pyrolysis at 411 and 424 °C, respectively,with the use of a superheated steam pyrolysis unit at Eprida Inc.(Athens, GA). Pinewood-derived biochar was produced frompinewood chips (that were cleaned and contained no bark orleaves) through a fast pyrolysis process at 450 °C using anauger reactor designed and fabricated at Auburn University.Further details on the production of these three biocharsamples can be found in Smith et al.12

Extraction of Biochar Water-Extractable Substances(BWES). The extraction of water-soluble substances frombiochar was performed by soaking 50 g of each biochar samplematerial (described above) in 200 mL of ultrahigh-quality(UHQ) Milli-Q water (Millipore). The biochar−water mixturewas placed on a shaker plate operating at 100 rpm for at least24 h. The liquid filtrates were transferred to 500 mL centrifugetubes and centrifuged at 5000 rpm (4682× g) with a JLA-10.5rotor at 4 °C for 15 min using a Beckman Coulter Avanti highspeed centrifuge (model no. J-26 XP). The supernatants andpellets were collected separately. Then, the collected super-natants were vacuum filtered through a filter using a 12 cmdiameter Buchner funnel and were then frozen and freeze-dried(170 Torr, −75 °C) to obtain the BWES as dry solids. The drymasses of the BWES were 1.693 g BWES per 50 g pinewood-derived biochar, 5.334 g BWES per 50 g chicken litter-derivedbiochar, and 0.247 g BWES per 50 g peanut shell-derivedbiochar. These BWES were redissolved in smaller amounts ofMilli-Q water in order to make stock solutions at the followingconcentrations: 28.2 g pinewood BWES per L, 28.2 g chickenlitter BWES per L, and 14.1 g peanut shell BWES per L (thislower concentration was due to the peanut shell BWES limiteddry mass). The ambient pH values of the pinewood BWES,chicken litter BWES, and peanut shell BWES stock solutionswere 3.94, 8.87, and 9.09, respectively. Prior to analysis usingESI-FTICR-MS, BWES stock solutions were desalted using amini-electrodialysis system (Harvard Apparatus) following theprocedures published previously for this method.12,16 For moredetailed information regarding the electrodialysis procedure,refer to the materials and methods section in the SupportingInformation (SI).

Dissolved Organic Carbon (DOC) Analysis. Sampleswere analyzed for their DOC concentrations by hightemperature catalytic combustion to CO2 on a ShimadzuTOC-VCPH total organic carbon analyzer, calibrated withpotassium hydrogen phthalate. The samples were acidifiedand sparged prior to DOC measurement, to remove Ccontributions from inorganic carbon.

Elemental and Proximal Analysis. Biochar samples weresent to Galbraith Laboratories (Knoxville, TN) for elementaland proximate analysis (modified ASTM D3172 and 3176). Formore information regarding the procedure, visit www.galbraith.com.

ESI-FTICR-MS Instrumentation and Analysis. Massspectral analysis was conducted with an Apollo II ESI ionsource on a Bruker Daltonics 12 T Apex Qe FTICR-MS in theCollege of Sciences Major Instrumentation Cluster (COSMIC)facility at Old Dominion University. All samples were analyzed

Environmental Science & Technology Article

dx.doi.org/10.1021/es4034777 | Environ. Sci. Technol. 2013, 47, 13294−1330213295

in negative ion mode, in which the ESI voltages, ionaccumulation, and number of coadded scans were optimized.All mass spectra were externally and internally calibratedaccording to previously published methods,17 while formulaassignment was done similar to Stubbins et al.18 Further detailsregarding sample preparation and ESI-FTICR-MS instrumentalparameters can be found in the Supporting Information.

■ RESULTS AND DISCUSSIONBiochar Elemental and Proximate Analysis. Table 1

shows the elemental and proximate results for the threedifferent biomass (pinewood, peanut shell, and chicken litter)derived biochar samples utilized here and in our previousstudy.12 The volatile matter (VM) content is widely used toassess the labile and stable components associated with biochar.The VM of both pinewood- and peanut shell-derived biocharwas around 40%, while chicken litter-derived biochar VM wassubstantially lower at 23%. The high percent ash and low fixedcarbon for chicken litter-derived biochar suggests the presenceof inorganic species that likely play a significant role in itscomposition. Supporting Information, Table S1 lists theelemental and proximate results of previous published data.

Kim et al.19 reported data for pinewood-derived biocharproduced by a single-auger pyrolysis system, similar to the oneused here, at roughly 450 °C that are comparable to thosefound in this study. Supporting Information, Table S1 also liststhe data for peanut shell- and chicken litter-derived biochararound 400 °C from Novak et al.20 and Lin et al.,21 respectively,which shows similar elemental compositions of the biocharsamples used in this study. This is a significant observation,since it indicates that biochars from different parts of the world,with similar elemental compositions, could also potentiallycontain toxic BWES, which should be carefully examined beforeconsidering large scale application.

Mass Spectral Characteristics of BWES. ESI-FTICRmass spectra for the ED separated fractions for pinewood- andpeanut shell-derived BWES are shown in SupportingInformation, Figure S1. In general, over 3000 peaks weredetected at 200−700 m/z for each of the biomass-derivedBWES. Like other complex mixtures (i.e., natural organicmatter), it is necessary to mention that the observed peaks areall singly charged, based on the observation of the 13C isotopepeak at 1.00335 m/z units (the mass difference of 13C and 12C)higher than the 12C parent peak (Supporting Information,

Table 1. Biochar Elemental (Oven-Dry wt % Basis) and Proximate Analysis from Galbraith Laboratories, Knoxville, TNa

biochar sample% loss ondrying % volatile matter @950 °C % ash

% fixed carbon (bydifference) C % H % N % O % S %

O:C(mol ratio)

pinewood biochar 4.6 38.9 1.6 57.6 73.5 3.4 <0.5 21.4 <0.05 0.22peanut shellbiochar

7.3 40.7 4.5 51.3 68.3 3.1 2.4 21.4 0.16 0.24

chicken litterbiochar

4.7 23.4 55.3 17.4 27.8 1.1 2.3 6.9 2.0 0.19

aAverage percentages of two analyses, with the exception of % loss on drying, which is from a one-time analysis. SD < 10%. SD = standard deviation.

Figure 1. van Krevelen diagrams (colored according to heteroatom content) for the molecular formulas assigned to mass spectral peaks inelectrodialysis-separated [anode-isolated (left), center chamber retained (middle), and cathode-isolated (right) substances] pinewood (top, a−c) andpeanut shell (bottom, d−f) derived biochar water-extractable substances (BWES). Boxes show molar H/C and O/C ratios of major biomolecularcompound classes (lipids, proteins, carbohydrates, unsaturated hydrocarbons, lignins, and condensed aromatics) outlined by Ohno et al.31

Environmental Science & Technology Article

dx.doi.org/10.1021/es4034777 | Environ. Sci. Technol. 2013, 47, 13294−1330213296

Figure S1a inset).22−24 Because of the low levels of organicsubstances, based on DOC measurements, and the biocharcomposition given in Table 1, chicken litter BWES wereomitted from FTICR-MS analysis after initial data showed onlythe detection of high mass defect (mass defect is defined as thedistance a peak is displaced from the exact nominal mass) peaksthat are likely of inorganic origin and could not be assignedmolecular formulas.25

The center-chamber retained substances of the pinewoodBWES (Supporting Information, Figure S1a) gave a differentpeak distribution when compared to the pinewood anode- andcathode-isolated substances that are more similar to each other(Figures S1a). Peanut shell anode- and cathode-isolatedsubstances (Figures S1b) both lacked the dense peakdistribution that the pinewood anode- and cathode-isolatedsubstances exhibited. For the pinewood-derived BWES in theanode- and cathode-isolates, the peaks with highest magnitudeswere detected at m/z 250−350, with smaller peak magnitudeclusters at m/z > 400. On the contrary, the center-chamberretained pinewood-derived BWES showed their greatestmagnitude between m/z 400−600, with smaller peak clustersabove and below this range. These observations are significantbecause they indicate that the anode- and cathode-isolatedsubstances likely contain smaller compounds (less than 500Da) as a result of the ED separation. On the other hand, thiswas not observed in the ED separation of peanut shell BWES.The greatest peak magnitudes for the center retainedsubstances were detected at m/z 250−350, with slightlylower peak magnitude clusters at m/z 400−550. Thisobservation, along with the fewer peaks detected in theanode- and cathode-isolated substances, suggests that thepeanut shell BWES most likely do not contain easily chargedspecies (i.e., carboxyl and hydroxyl groups) that are character-istic of the hypothesized inhibiting compound(s) associatedwith pinewood BWES.Molecular Formula Characteristics of the BWES. To

compare the assigned molecular formulas of the ED separatedpinewood BWES and peanut shell BWES fractions, vanKrevelen diagrams (Figure 1) are utilized to highlight similarcharacteristics of common types of biomolecular compoundclasses (i.e., proteins, carbohydrates, lignin, etc.) tentativelyassigned to the formulas based on their molar H/C (y-axis) andmolar O/C (x-axis) ratios. These plots have been used in thepast to categorize varying compound classes present inbiochar,14 dissolved organic matter,26−28 black carbon-derivedDOM,29 aerosol-derived water-soluble organic carbon,30 andplant biomass-derived organic matter.31

Figure 1 panels a−f show van Krevelen plots for themolecular formulas identified in each of the 3 ED separatedfractions [anode-isolated substances (left, a and d), centerchamber retained substances (middle, b and e), and cathode-isolated substances (right, c and f)] for pinewood (top) andpeanut shell (bottom) BWES analyzed by ESI-FTICR-MS. It isapparent that both pinewood (top) and peanut shell (bottom)BWES contain many formulas with H/C and O/C ratioscharacteristic of lignin (O/C = 0.1−0.67 and H/C = 0.7−1.5).Figure 1 also highlights that both BWES show nitrogen, sulfur,and phosphorus containing formulas that encompass a varietyof major classes of organic compounds, shown by the overlainboxes. These findings are consistent with those recentlypublished by Lin et al.13 on water extractable organic carbonin untreated and chemically treated biochars characterized byliquid chromatography−organic carbon detection. Although

pinewood and peanut shell BWES both contain molecularformulas predominantly characteristic of lignin-type species,they have differences as well. Pinewood BWES contains moreformulas with high O/C and H/C ratios, characteristic ofcarbohydrate-like compounds, when compared to peanut shellBWES. Likewise, peanut shell BWES contain a significantlylarger number of formulas with low O/C and H/C ratios,indicating a more aromatic nature. In both cases, it is obviousthat BWES are highly complex in nature and their character-istics are examined in more detail below.

Possible Inhibiting Compound(s) Associated withPinewood BWES. Our previous study showed that pinewoodBWES contained an estimated 56.2% DOC in the total soluteand exhibited an acidic pH (3.94), whereas peanut shell andchicken litter BWES contained significantly lower amounts ofDOC (10.7% and 1.0%) and more basic pH values (8.87 and9.09, respectively).12 It was determined through toxicologicalbioassays with pH-neutralized and DOC-normalized pinewoodBWES that the amount of DOC and acidic nature of pinewoodBWES did not contribute to inhibition of both prokaryotic(Synechococcus sp.) and eukaryotic (Desmodesmus sp.) algalgrowth, but that certain specific component(s) within thepinewood BWES caused the growth inhibition. We alsodemonstrated that, upon ED separation of pinewood BWES,the toxic compound(s) of pinewood BWES were predom-inantly isolated in the anode chamber, where negatively chargedcompounds (anions) most likely reside. This observationsuggests that the inhibitory compounds most likely containnegatively charged functional groups, such as a carboxyl group,that enables the use of negative ion mode ESI-FTICR-MS as aneffective characterization technique. Figures 1a and 1c showthat more than 1700 molecular formulas were each assigned tothe anode- and cathode-isolated substances in pinewoodBWES, while less than a combined total of 700 formulaswere assigned to the anode- and cathode-isolated substances inpeanut shell BWES (Figures 1d,f). The van Krevelen diagramscorrelate well with the observed peak distribution discussedabove, which also shows that peanut shell-derived BWEScontain a significant amount of smaller molecular weight peaks(below 500 Da, Supporting Information, Figure S1b). Becausevery few peaks are detected during FTICR-MS analysis of theanode- and cathode-isolated substances for peanut shell-derivedBWES (Figure 1d,f), it is likely that they do not containfunctional groups that become easily charged in water (duringED), assuming that this is not due to matrix effects or ionsuppression. This observation is significant because ourprevious study demonstrated that the proposed inhibitingcompound(s) in the pinewood BWES readily passed throughthe ED membranes, with a substantially higher amount of toxiccompound(s) located in the anode-isolated substances.12 Wenote that in Figure 1d, peanut shell BWES anode-isolatedsubstances, display a cluster of N-containing formulas with H/C and O/C ratios characteristic of peptides, indicating thatduring ED these types of compounds will migrate to the anode.To identify the compound(s) that possibly inhibit algae

growth, an in-depth comparison of the pinewood BWES anode-isolated, cathode-isolated, and center chamber retainedsubstances was conducted. Our previous study showed thatall pinewood BWES ED separated fractions demonstrated somedegree of inhibition on the growth of blue-green algae(Synechococcus), when compared to the control, with theanode-isolated substances demonstrating the highest toxicity onalgal growth.12 Figure 2a shows the van Krevelen diagram of

Environmental Science & Technology Article

dx.doi.org/10.1021/es4034777 | Environ. Sci. Technol. 2013, 47, 13294−1330213297

the molecular formulas that are common to all pinewoodBWES ED separated fractions, where it is highly possible thatthe inhibiting compound(s) reside. This plot suggests that thepinewood inhibiting species resemble lignin-type compoundsthat likely contain carboxyl and hydroxyl functionalities. Whena DC voltage is applied to the ED system, compounds withnegatively charged functional groups, such as a deprotonatedcarboxyl group, will migrate toward the anode chamber, whilesome of them may stay charge neutral and reside in the centerchamber. Certain negatively charged organic species can alsoundergo chelation with high valent metal cations, such as Fe2+,to form a chelated organic acid−metal complex with netpositive charge, [R-COO Fe]+, and thus migrate toward thecathode chamber. It is important to point out that due to themolecular size cutoff of the cellulose membrane (500 Da),compounds below 500 Da may also diffuse into the anode andcathode chambers based on size alone. However, this cannotexplain the reason for a significantly higher amount of toxicitybeing observed for the anode-isolated substances. Because thedistribution pattern in Figure 2a only occupies a few of thebiomolecular regions outlined by Ohno et al.31 (boxes overlainin Figure 1), we have eliminated all but two regions, lignin andcarbohydrate, to aid in the investigation of the inhibitingcompound(s). About 70% of the formulas common to all threeseparated fractions fall within the lignin region. The vastmajority (98%) of these formulas are CHO-only, with nearlynegligible contributions from CHOS and CHON formulas.While the heteroatom content of the formulas found in all threefractions is minor, it is important to mention that these types of

formulas (and other unique molecular formulas) are detectedpredominantly in the anode- and cathode-isolated substances,especially if they contain charged functional groups that allowedthe compounds to move freely during the ED process.In the analysis above, we assumed that the molecular

formulas common to all pinewood BWES ED separatedfractions contain the toxin that caused the inhibition of algalgrowth. The other possibility is that the toxic species would notbe common to all pinewood BWES ED separated fractions.Therefore, anode-isolated pinewood BWES would contain themost potent toxic species while the cathode-isolated and centerchamber retained BWES could contain different but less toxicspecies. In that case, the specific molecular formulas found inthe anode-isolated substances can be qualitatively comparedwith those in the cathode-isolated and center chamber retainedsubstances. Figure 2b−d shows the van Krevelen diagrams offormulas assigned to the anode-isolated (2b, left), centerchamber retained (2c, middle), and cathode-isolated (2d, right)substances that were not detected in all three pinewood BWESseparated fractions (i.e., subtraction of Figure 2a from Figure 1panels a, b, and c). It is important to note that the formulasshown in Figure 2b−d are not necessarily unique to eachfraction, because the formula could be detected in two of thethree fractions. When the van Krevelen plot in Figure 2b(pinewood BWES anode-isolated substances) is compared tothat in Figure 2c (pinewood BWES center chamber retainedsubstances), a large number of formulas that contain CHO-onlyand a large cluster of CHON compounds in the lignin-likeregion are observed in the center chamber retained substances.

Figure 2. van Krevelen diagrams that plot the molecular formulas (colored according to heteroatom content) that are (a) common to all threeelectrodialysis-separated fractions (anode-isolated, center chamber retained, and cathode-isolated substances) of pinewood-derived BWES, and thosethat are (b) found in the anode-isolated, (c) center chamber retained, and (d) cathode-isolated fractions, after removing those that are common to allthree. Boxes show molar H/C and O/C ratios of the lignin and carbohydrate regions outlined by Ohno et al.31

Environmental Science & Technology Article

dx.doi.org/10.1021/es4034777 | Environ. Sci. Technol. 2013, 47, 13294−1330213298

Because these formulas predominantly exist in the centerchamber retained substances and not the anode-isolatedsubstances, it is unlikely that they are representative of theinhibiting compounds. A unique group of formulas detectedonly in the pinewood BWES anode-isolated substances is thelarge cluster of CHON compounds located in the carbohydrateregion. This group of formulas is absent in the cathode-isolatedsubstances and the center chamber retained substances,suggesting that these formulas are also unlikely to representthe inhibiting species, since our previous study12 showed thatcathode-isolated substances and the center chamber retainedsubstances still show minor toxicity. However, this conclusion isbased on the assumption that algal growth inhibition is due toone type of compound as opposed to several different types ofcompounds.After eliminating the most unlikely inhibitory groups of

formulas, three notable clusters of compounds still remain: thecondensed CHOS formulas that contain one sulfur atom in thelignin region, the aliphatic CHO formulas located where thelignin and carbohydrate regions overlap, and the phosphoruscontaining formulas (Figure 2b). Figure 2b−d highlights theabundance of these groups of formulas in the anode-isolatedsubstances, with fewer assigned to the cathode-isolated andcenter chamber retained substances. To compare the number ofassigned formulas for the three electrodialysis separatedfractions, Venn diagrams are utilized to show the overlap offormulas found in each fraction. Figure 3 highlights thecondensed CHOS (Figure 3a), aliphatic CHO (Figure 3b), and

phosphorus containing (Figure 3c) formulas that are possibleinhibitor species. All Venn diagrams in Figure 3 show that theanode-isolated substances contain the highest percentage ofunique molecular formulas (i.e., those that do not overlap withthe other two samples). This observation is important becauseinhibition is significantly higher for the anode-isolatedsubstances, suggesting that more compounds similar in natureto the inhibiting species should be found in the anode-isolatedsubstances. A major difference in the Venn diagrams of Figure 3is the percentage of formulas assigned to all three fractions.Both the sulfur and phosphorus containing formulas show ≤2%of the formulas were assigned to all three fractions, but that asignificant amount of each are common to both the anode- andcathode-isolated substances (Figure 3a and 3c). Thisobservation correlates well with the bioassay experimentspreviously published,12 where the cathode-isolated substancesshowed more inhibition then the center chamber retainedsubstances. The aliphatic CHO formulas show an oppositetrend, in which 67% of the formulas were assigned to all threefractions, with another 14% of formulas common to only theanode-isolated and center chamber retained substances. Thisobservation suggests that a number of similar compounds existin all three of the ED separated fractions, where their actualconcentration could play a vital role in algal growth inhibition.Because ESI-FTICR-MS is not a quantitative technique, we canonly comment here on the number of formulas assigned to thevarious fractions and not speculate on how that could relate totheir concentration within each fraction.

Kendrick Mass Defect Analysis. To more closely examinemolecular formulas and how they are related to one another,many have found it useful to perform Kendrick mass defect(KMD) analysis.32−34 KMD analysis is often utilized to findand highlight formulas that fall into homologous series basedon m/z values that differ only by the exact mass of a certainfunctional group (i.e., a CH2 group). After a particularfunctional group is chosen, the Kendrick mass of a molecularformula is calculated by multiplying the ratio of the nominalmass to the exact mass (of the functional group) by thecalculated exact m/z of the assigned formula. The KMD is thencalculated by subtracting the Kendrick mass from the nominalmass. This KMD value can be plotted against the nominalKendrick mass, to form what is known as a KMD plot. In thisplot, homologous series fall on the same horizontal line,because the formulas have identical KMD values and differ bythe nominal mass of the functional group. We hypothesize thatat least one carboxyl group is a specific characteristic of theinhibitor and could most likely account for the majority of theinhibiting compound(s) migration to the anode-isolatedsubstances and contribute to the acidic pH (3.94) associatedwith the original pinewood BWES. Figure 4a−c shows theKMD plots of carboxyl (COO) groups for the commonmolecular formulas in all three pinewood BWES fractions(Figure 4a), condensed CHOS formulas (Figure 4b), andaliphatic CHO formulas (Figure 4c). Because of the smallnumber of phosphorus containing formulas, the KMD analysisis not shown as it did not exhibit any series containing morethan two members for any KMD value. SupportingInformation, Figure S2 shows an expanded region of theKMD plots shown in Figure 4 panels a−c, to highlight thenumerous series found when KMD analysis is employed. Figure4a−c highlights that in all three groups, numerous series offormulas that differ only by a carboxyl group (ranging from 2 to5 carboxyl groups) are present. This finding suggests that it is

Figure 3. The Venn diagrams of three groups of possible inhibitorycompounds (based on molecular formula assignment). Percentages inareas of overlap are molecular formulas that appear in both (or allthree) of those samples. Percentages in areas with no overlap aremolecular formulas that are unique to that individual sample.

Environmental Science & Technology Article

dx.doi.org/10.1021/es4034777 | Environ. Sci. Technol. 2013, 47, 13294−1330213299

highly likely that the inhibiting species contains carboxylfunctionality and could explain both their migration into theanode chamber during ED and contribute to the acidic pH(3.94) of pinewood BWES.Supporting Information, Figure S2a−d shows KMD plots for

water (left) and methoxyl (right) for the common molecularformulas in all three pinewood BWES fractions (Figure S2a),condensed CHOS formulas (Figure S2b), aliphatic CHOformulas (Figure S2c), and phosphorus containing formulas(Figure S2d). These plots show that not only do these possibleinhibitors most likely contain carboxyl groups, they also allshow hydroxyl and methoxyl functionality as well. All threefunctionalities highlighted by KMD analysis are commonlyassociated with functionality of degraded lignin, and thus, areworth discussing.

Lignins are complex, high-molecular-weight biopolymerscomposed of phenylpropanoids primarily linked by ether-linkages. Lignin is formed from the enzyme-initiated randomfree-radical copolymerization of three different phenylpropanemonomers (p-coumaryl alcohol, coniferyl alcohol, and sinapylalcohol), often referred to as monolignols.35 Its compositionvaries in different plants, but lignin is most commonly found inwood cells. It is known that softwoods, (i.e., pinewood) containthe highest lignin content, around 30%,36 followed byhardwoods and grasses. Because of its complex nature, muchis still unclear about lignin’s exact structure, and thus the abilityto fully characterize the pyrolytic degradation is limited. Aspreviously mentioned, Lin et al.13 found a variety of waterextractable organic carbon substances from multiple biomasspyrolyzed biochars, characteristic of high molecular weightbiopolymers, humic substances, and other components usingLC-OCD. During pyrolysis, lignin-containing biomass under-goes four stages of thermal degradation: moisture evolution,hemicellulose decomposition, cellulose decomposition, andlignin decomposition.37 Previous studies show that decom-position of hemicellulose and cellulose occur below 400 °C,14,36

while the decomposition of lignin can vary as high as 900 °C,according to Yang et al.37 In this study, pinewood biomass waspyrolyzed at 450 °C, and thus thermal decomposition ofhemicellulose and cellulose is likely to have occurred. We haveshown using ESI-FTICR-MS that molecular formulas assignedto BWES contain O/C and H/C ratios that largely resemblethose characteristic of degraded lignin, with minor contribu-tions from hemicellulosic and cellulosic derived species. Thevan Krevelen plots reveal four groups of possible inhibitorspecies (based on molecular formulas) ranging from low O/Cand low H/C (characteristic of condensed aromatics) to highO/C and high H/C (characteristic of more aliphaticcompounds), which were further subjected to KMD analysis.KMD analysis strengthened our previous hypothesis that theinhibiting species contains carboxyl functionality, when threepossible inhibitory groups of formulas with varying O/C andH/C ratios as possible candidates are used. KMD analysis alsoshowed that the possible inhibitor species most likely containhydroxyl and methoxyl functionalities commonly associatedwith the pyrolytic degradation of lignin. Although KMD andvan Krevelen analysis from FTICR-MS data can providevaluable information regarding functionality, structural in-formation is not available solely on the basis of the knowledgeof a molecular formula. Further isolation of the inhibitor speciesis currently being explored, so that subsequent FTICR-MSanalysis can be utilized as a critical tool for the analysis ofBWES.In this study, we have shown the effectiveness of ultrahigh

resolution mass spectrometry and its application for under-standing the relationship between biochar water extractablesubstances and the inhibition of aquatic photosyntheticmicroorganisms growth, at a molecular level. We revealedthat growth inhibition of aquatic photosynthetic microorgan-isms is most likely due to degraded lignin-like species rich inoxygen containing functionalities. The new application ofFTICR-MS presented in this study has shown potential to beutilized as a valuable characterization technique for studying thepossible contaminants that would leach into the aquaticecosystems if biochar were used as both a carbon sequestrationtechnique and soil amendment on the scale of gigatons ofcarbon.

Figure 4. Kendrick mass defect plots using the carboxyl functionalityfor (a) the molecular formulas assigned to all three electrodialysis-separated pinewood BWES fractions (molecular formulas shown inFigure 2a), (b) condensed CHOS formulas (Figure 3a), and (c)aliphatic CHO formulas (Figure 3b). Expanded regions of each plotare shown in the Supporting Information (Figure S2).

Environmental Science & Technology Article

dx.doi.org/10.1021/es4034777 | Environ. Sci. Technol. 2013, 47, 13294−1330213300

■ ASSOCIATED CONTENT*S Supporting InformationExpanded materials and methods section, as well as the figuresreferenced in the manuscript. This material is available free ofcharge via the Internet at http://pubs.acs.org.

■ AUTHOR INFORMATIONCorresponding Author*E-mail: jwlee@odu.edu.NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSWe thank Dr. Sandeep Kumar and Mr. Danny Day forproviding the biochar samples used in this study. We thankSusan Hatcher and Jared Callan at the COSMIC (College ofSciences Major Instrumentation Cluster) facility at ODU forFTICR-MS assistance. This research was supported, in part, bythe 2013 Old Dominion University Multidisciplinary SeedFunding Program, and jointly by Dr. Hatcher’s ODU FrankBatten Endowment funds and by Dr. Lee’s start-up researchfunds provided by the Department of Chemistry andBiochemistry, the College of Sciences, the Office of Researchat ODU, and the ODU Research Foundation. Also, we thankAmanda Willoughby and Wassim Obeid for their assistance andguidance in mass spectral data analysis.

■ REFERENCES(1) Day, M. D.; Evans, R. J.; Lee, J. W.; Reicosky, D. C. EconomicalCO2, SOx, and NOx capture from fossil-fuel utilization with combinedrenewable hydrogen production and large-scale carbon sequestration.Energy 2005, 30, 2558−2579.(2) Downie, A. E.; Zwieten, L. V.; Smernik, R. J.; Morris, S.; Munroe,P. R. Terra Preta Australis: Reassessing the carbon storage capacity oftemperate soils. Agr. Ecosyst. Environ. 2011, 140, 137−147.(3) Krausmann, F.; Erb, K.; Gingrich, S.; Lauk, C.; Haberl, H. Globalpatterns of socioeconomic biomass flows in the year 2000: Acomprehensive assessment of supply, consumption and constraints.Ecol. Econ. 2008, 65, 471−487.(4) Woolf, D.; Amonette, J. E.; Street-Perrott, F. A.; Lehmann, J.;Joseph, S. Sustainable biochar to mitigate global climate change. Nat.Commun. 2010, 1, 56 DOI: 10.1038/ncomms1053.(5) Lee, J. W.; Hawkins, B.; Day, D. M.; Reicosky, D. C.Sustainability: The capacity of smokeless biomass pyrolysis for energyproduction, global carbon capture and sequestration. Energ. Environ.Sci. 2010, 3, 1609−1812.(6) Rondon, M. A.; Lehmann, J.; Ramírez, J.; Hurtado, M. Biologicalnitrogen fixation by common beans (Phaseolus vulgaris L.) increaseswith bio-char additions. Biol. Fertil. Soils 2007, 43, 699−708.(7) Gundale, M. J.; Thomas, H.; DeLuca, T. H. Charcoal effects onsoil solution chemistry and growth of Koeleria macrantha in theponderosa pine/Douglas fir ecosystem. Biol. Fertil. Soils 2007, 43,303−311.(8) Deenik, J. L.; McClellan, T.; Uehara, G.; Antal, M. J.; Campbell,S. Charcoal volatile matter content influences plant growth and soilnitrogen transformations. Soil Sci. Soc. Am. J. 2010, 74, 1259−1270.(9) Spokas, K. A.; Baker, J. M.; Reicosky, D. C. Ethylene: potentialkey for biochar amendment impacts. Plant Soil 2010, 333, 443−452,DOI: 10.1007/s11104-010-0359-5.(10) Olsson, M.; Ramnas, O.; Petersson, G. Specific volatilehydrocarbons in smoke from oxidative pyrolysis of softwood pellets.J. Anal. Appl. Pyrol. 2004, 71, 847−854.(11) Song, J. Z.; Peng, P. A. Characterization of black carbonmaterials by pyrolysis−gas chromatography−mass spectrometry. J.Anal. Appl. Pyrol. 2010, 87, 129−137.

(12) Smith, C. R.; Buzan, E. M.; Lee, J. W. Potential impact ofbiochar water-extractable substances on environmental sustainability.ACS Sustainable Chem. Eng. 2013, 1, 118−126.(13) Lin, Y.; Munroe, P.; Joseph, S.; Henderson, R.; Ziolkowski, A.Water extractable organic carbon in untreated and chemical treatedbiochars. Chemosphere 2012, 87, 151−157.(14) Podgorski, D. C.; Hamdan, R.; McKenna, A. M.; Nyadong, L.;Rodgers, R. P.; Marshall, A. G.; Cooper, W. T. Characterization ofpyrogenic black carbon by desorption atmospheric pressure photo-ionization Fourier transform ion cyclotron resonance mass spectrom-etry. Anal. Chem. 2012, 84, 1281−1287.(15) Cole, D. P.; Smith, E. A.; Lee, Y. J. High-resolution massspectrometric characterization of molecules on biochar from pyrolysisand gasification of switchgrass. Energy Fuels 2012, 26, 3803−3809.(16) Chen, H.; Stubbins, A.; Hatcher, P. G. A mini-electrodialysissystem for desalting small volume saline samples for Fourier transformion cyclotron resonance mass spectrometry. Limnol. Oceanogr.-Meth.2011, 9, 582−592.(17) Sleighter, R. L.; McKee, G. A.; Liu, Z.; Hatcher, P. G. Naturallypresent fatty acids as internal calibrants for Fourier transform massspectra of dissolved organic matter. Limnol. Oceanogr.-Meth. 2008, 6,246−253.(18) Stubbins, A.; Spencer, R. G. M.; Chen, H.; Hatcher, P. G.;Mopper, K.; Hernes, P. J.; Mwamba, V. L.; Mangangu, A. M.;Wabakanghanzi, J. N.; Six, J. Illuminated darkness: Molecularsignatures of Congo River dissolved organic matter and itsphotochemical alteration as revealed by ultrahigh precision massspectrometry. Limnol. Oceanogr. 2010, 55, 1467−1477.(19) Kim, P.; Johnson, A.; Edmunds, C. W.; Radosevich, M.; Vogt,F.; Rials, T. G.; Labbe, N. Surface functionality and carbon structuresin lignocellulosic-derived biochars produced by fast pyrolysis. EnergyFuels 2011, 25, 4693−4703.(20) Novak, J. M.; Lima, I.; Xing, B.; Gaskin, J. W.; Steiner, C.; Das,K. C.; Ahmedna, M.; Rehrah, D.; Watts, D. W.; Busscher, W. J.;Schomberg, H. Characterization of designer biochar produced atdifferent temperatures and their effects on a loamy sand. Ann. Environ.Sci. 2009, 3, 195−206.(21) Lin, Y.; Munroe, P.; Joseph, S.; Ziolkowski, A.; Zwieten, L.;Kimber, S.; Rust, J. Chemical and structural analysis of enhancedbiochars: thermally treated mixtures of biochar, chicken litter, clay andminerals. Chemosphere 2013, 91, 35−40.(22) Kujawinski, E. B.; Freitas, M. A.; Zang, X.; Hatcher, P. G.;Green-Church, K. B.; Jones, R. B. The application of electrosprayionization mass spectrometry (ESI MS) to the structural character-ization of soil organic matter. Org. Geochem. 2002, 33, 171−180.(23) Stenson, A. C.; Landing, W. M.; Marshall, A. G.; Cooper, W. T.Ionization and fragmentation of humic substances in electrosprayionization Fourier transform-ion cyclotron resonance mass spectrom-etry. Anal. Chem. 2002, 74, 4397−4409.(24) Kim, S.; Kramer, R. W.; Hatcher, P. G. Graphical method foranalysis of ultrahigh-resolution broadband mass spectra of naturalorganic matter, the Van Krevelen diagram. Anal. Chem. 2003, 75,5336−5344.(25) Sleighter, R. L.; Hatcher, P. G. Fourier transform massspectrometry for the molecular level characterization of natural organicmatter: Instrument capabilities, applications, and limitations. In FourierTransformsApproach to Scientific Principles; Nikolic, G., Ed.; InTech:Rijeka, Croatia, 2011; ISBN: 978-953-307-231-9; http://www.intechopen.com/articles/show/title/fourier-transform-mass-spectrometry-for-the-molecular-level-characterization-of-natural-organic-matt.(26) Kim, S.; Simpson, A. J.; Kujawinski, E. B.; Freitas, M. A.;Hatcher, P. G. Non-invasive advanced spectroscopic methods(electrospray ionization mass spectrometry and 2D solution NMR)for analysis of DOM isolated by C18 solid phase disk extraction. Org.Geochem. 2003, 34, 1325−1335.(27) Kujawinski, E. B.; Vecchio, R. D.; Blough, N. V.; Klein, G. C.;Marshall, A. G. Probing molecular-level transformations of dissolvedorganic matter: insights on photochemical degradation and protozoan

Environmental Science & Technology Article

dx.doi.org/10.1021/es4034777 | Environ. Sci. Technol. 2013, 47, 13294−1330213301

modification of DOM from electrospray ionization Fourier transformion cyclotron resonance mass spectrometry. Mar. Chem. 2004, 92, 23−37.(28) Sleighter, R. L.; Hatcher, P. G. Molecular characterization ofdissolved organic matter (DOM) along a river to ocean transect of thelower Chesapeake Bay by ultrahigh resolution electrospray ionizationFourier transform ion cyclotron resonance mass spectrometry. Mar.Chem. 2008, 110, 140−152.(29) Hockaday, W. C.; Grannas, A. M.; Kim, S.; Hatcher, P. G. Directmolecular evidence for the degradation and mobility of black carbon insoils from the ultrahigh-resolution mass spectral analysis of dissolvedorganic matter from a fire-impacted forest soil. Org. Geochem. 2006,37, 501−510.(30) Wozniak, A. S.; Bauer, J. E.; Sleighter, R. L.; Dickhut, R. M.;Hatcher, P. G. Technical Note: Molecular characterization of aerosol-derived water soluble organic carbon using ultrahigh resolutionelectrospray ionization Fourier transform ion cyclotron resonancemass spectrometry. Atmos. Chem. Phys. 2008, 8, 5099−5111.(31) Ohno, T.; He, Z.; Sleighter, R. L.; Honeycutt, C. W.; Hatcher, P.G. Ultrahigh resolution mass spectrometry and indicator speciesanalysis to identify marker components of soil- and plant biomass-derived organic matter fractions. Environ. Sci. Technol. 2010, 44, 8594−8600.(32) Kendrick, E. A mass scale based on CH2 = 14.0000 for highresolution mass spectrometry of organic compounds. Anal. Chem.1963, 35, 2146−2154.(33) Kramer, R. W.; Kujawinski, E. B.; Hatcher, P. G. Identificationof black carbon derived structures in a volcanic ash soil humic acid byFourier transform ion cyclotron resonance mass spectrometry. Environ.Sci. Technol. 2004, 38, 3387−3395.(34) Sleighter, R. L.; Hatcher, P. G. The application of electrosprayionization coupled to ultrahigh resolution mass spectrometry for themolecular characterization of natural organic matter. J. Mass Spectrom.2007, 42, 559−574.(35) Klein, M. T.; Virk, P. S. Modeling of lignin thermolysis. EnergyFuels 2008, 22, 2175−2182.(36) Liu, Q.; Wang, S.; Zheng, Y.; Luo, Z.; Cen, K. Mechanism studyof wood lignin pyrolysis by using TG-FTIR analysis. J. Anal. Appl.Pyrolysis 2008, 82, 170−177.(37) Yang, H.; Yan, R.; Chen, H.; Lee, D. H.; Zheng, C.Characteristics of hemicellulose, cellulose and lignin pyrolysis. Fuel2007, 86, 1781−1788.

Environmental Science & Technology Article

dx.doi.org/10.1021/es4034777 | Environ. Sci. Technol. 2013, 47, 13294−1330213302