The chemical industry is still heavily reliant on the use of fossil raw materials and a great need for alternative, sustainable ones exists. Lignocellulosic biomasses have received much scientific interest as renewable substitutes for different chemical industry applications, and thus the need for faster production of woody biomass is growing globally. This topic is timely also in Finland, where the rotation periods of commercially important tree species exceed over 70–80 years. One possible way to boost the production and availability of wood-based raw materials is to breed fast growing tree species designed for the chemical industry’s application purposes.
Hybrid aspen (Populus tremula
L. × tremuloides
Michx.) is one of the fastest growing tree species in Finland. On fertile sites in southern Finland, the yields may reach up to 20 m3
per year in about 25-year rotations [1
]. The species show excellent coppice vigor that enables even higher second rotation growth [2
]. High-yield capacity and chemical properties of hybrid aspen indicate its promising potential as raw material for biochemicals and green energy [3
]. Hybrid aspen could be an alternative for Norway spruce, which is presently predominantly used in reforestation of vigorous sites in Finland, but which faces severe problems due to Heterobasidion
root rot disease [5
] as well as with the spruce bark beetle (Ips acuminatus
], the risks of which are expected to increase with climate change.
Currently, bark of coniferous species is a major industrial by-product in the Nordic countries, being one of the most promising resource for added-value biochemical production. In Finland alone, the processes of forest industries use ca. 71 million m3
of round wood annually [7
]. The amount of bark is approximately 10% of the round wood volume; thus, the ca. 7 million m3
of bark is produced as a by-product every year. This residue is still mainly used for energy production.
Like wood, aspen bark mainly consists of lignin, hemicellulose and cellulose, and earlier studies have shown that the bark of Populus
family) are especially rich in extractable biochemical components. However, these phytochemicals have been poorly characterized [8
]. Only a few have been extensively explored as pharmaceuticals (e.g., salicylic and benzoate-based drugs) and reached commercialization. Aspen bark extractives have potential to be utilized as natural source of antioxidants. Populus tremuloides
Michx. bark hot water extract has reported to have greater antioxidative capacity than synthetic antioxidant butylated hydroxytoluene (BHT). With fractionating the bark crude water extracts with organic solvents, products antioxidative nature can be enhanced [9
]. The lignin content has been reported to reach ca. 15% and the rest is composed of carbohydrate-derived compounds [10
]. Additionally, the lipid content of aspen bark is 10% [11
], but the lipid composition is still undetermined. The suberin (i.e., suberic acids) content of extractive free Populus tremula
bark has been shown to be 37.9% [12
]. Based on the existing literature, aspen bark is thus a very promising, still heavily under-utilized raw material.
Material science has great interest to develop environmentally friendly composite materials to substitute petroleum-based polymers, in which lignocellulosic biomasses have characteristics to be utilized in such applications. Aspen tree fiber material has been studied as filler in biocomposite and plastic-wood composite matrices. With designed compositions of polymeric ingredients, aspen tree materials showed potential to be used as part of composite materials [13
For finding new value-added uses, more detailed clonal screening of the chemical diversity of bark is needed for aspen and poplar species and/or clones well adapted in the Finnish climate. This study aims to promote comprehensive utilization of woody biomass by providing a knowledgebase on aspen bark as a new alternative source for fossil-based chemicals. The research topics included analysis of clonal variation in: (1) major chemical components, i.e., hemicelluloses, cellulose, and lignin; (2) extraneous materials, i.e., bark extractives, and suberic acid; (3) condensed tannins content and composition; and (4) screening differences in antioxidative properties and total phenolic content of hot water extracts and ethanol-water extracts of hybrid aspen bark. For testing the hypothesis that no marked clonal differences in bark chemical properties exist, hybrid aspen trees representing three different clones with good growth and quality characteristics (based on earlier studies) were harvested, and bark samples at heights of 1.3 m and 5 m on the stem were analyzed.
4. Materials and Methods
4.1. Sample Trees
Three trees from each of three hybrid aspen (Populus tremula
L. × P. tremuloides
Michx.), clones (no. 2, 4 and 5) were randomly selected from a random-block designed field trial of the Natural Resource Institute Finland (Luke) in Punkaharju, in southeastern Finland. For each clone each tree was selected from a different block. The trial was established in May 2002 with 3 × 3 m plant spacing and the sample trees were felled in February 2018. For this study, from each tree two sample discs were collected from height 1.3 m and 5 m. Characteristics of the sample trees are presented in Table 1
4.2. Statistical Analysis
All dependent variables were analyzed by generalized linear model (GLM) having clone (2, 3 and 4), height (1.3 m and 5 m), and their interaction as fixed effects. Three replicates were used leading to sample size of 18 per variable. Correlation between heights within same experimental unit was taken into account using a heterogeneous or a homogeneous compound symmetry structure. The first one allows unequal variances for both heights and the latter handles those as equal, respectively. The most suitable information criterion (AICc) for small sample size was used in comparison of the covariance structures, and a likelihood test was used as a side to help the decision making [39
]. The normality of residuals was studied by graphically from multiple residual plots and founded adequate.
Only lignin, hemicellulose and ash were analyzed with the assumption of normal distribution. GLM with the assumption of lognormal (with an identity link) or Gamma distribution (with a log link) was used, when distributions of dependent variables were highly skewed. The assumption of beta distribution (with a logit link) was used for percentages. Restricted maximum likelihood estimation method (REML) was used for normal and lognormal distributions and residual pseudo-likelihood estimation method (REPL) for gamma and beta distributions, respectively [40
Six cross-comparisons (e.g., clone 2 in height 1.3 m vs. clone 4 in height 5 m) of the interaction term were excluded to minimize the amount of pairwise comparisons. The stepdown method of Westfall [41
] was used for pairwise comparisons of means (significance level of α = 0.05), which is known to be one of the most effective in cases when the design is balanced. Kenward–Roger method [42
] was used for calculating the degrees of freedom.
The analyses were performed using the GLIMMIX procedure of the SAS Enterprise Guide 7.15 (SAS Institute Inc., Cary, NC, USA).
Differences in the chemical properties between stem heights and clones were calculated as percentage difference (Equation (1)). Compared values are expressed in Figure 1
, Figure 2
, Figure 3
, Figure 4
and Figure 5
4.3. Wood Analysis
4.3.1. Sample Preparation
Bark samples for chemical analysis were manually separated from freshly cut aspen tree trunks. Bark mass contained both inner and outer bark layers. Bark materials were first freeze-dried and then grinded by a Pulverisette cutting mill (Type 15.903, Fritsch, Idar-Oberstein, Germany) with 2 mm sieve cassette. Powder with particle-size of <2 mm was used for cellulose, hemicellulose, lignin and suberic acid content analysis. First, samples were extracted with accelerated solvent extraction (ASE-350, Dionex, Sunnyvale, CA, USA) method in order to remove lipophilic and hydrophilic extractives. To increase the effectiveness of lignin and suberic acid extraction, extract free samples were further milled into fine powder with an IKA A 10 analytical mill (Kinematica, Littau/Luzern, Switzerland) equipped with a cooling unit (−20 °C).
Fine bark powder for the assays of antioxidative properties
and analysis of total phenol content
and condensed tannin
were prepared by first grinding the samples by using a Fritsch Pulverisette mill and then by an A 10 universal batch mill (IKA, Staufen, Germany). Powdered samples’ particle-size was estimated with test sieve set of 1.25, 0.5, 0.2, and 0.0063 mm sieves from one bark sample after IKA A 10 grind. Particle-size distribution estimates are expressed in Table 2
. Moisture content of the samples was checked with moisture analyzer (Moisture analyzer MLB 50-3N, KERN & Sohn GmbH, Balingen, Germany).
4.3.2. Lipohilic and Hydrophilic Extractives
Lipophilic and hydrophilic extractives was gravimetrically analyzed with accelerated solvent extraction (Dionex ASE-350) method. For extractions, bark samples were loaded to stainless steel cylinders weighing approximately 7.5–8.5 g each. Bark samples lipophilic extractives was extracted with hexane 90 °C, 3 × 15 min cycles and hydrophilic extractives with 95% acetone (aq) 100 °C, 3 × 15 min cycles. Each lipophilic and hydrophilic extractives end volume was adjusted to 50 mL with extraction solvents. Extractive content was determinated gravimetrically from 6 mL of extract by drying it under nitrogen gas flow in 40 °C temperature-controlled bath and measuring the weight of dry solid residues.
4.3.3. Suberic Acids
Suberic acids were determined from 2 g of extract-free samples. Samples were weighed out into 50 mL seal tight test tubes and 25 mL of 3% potassium hydroxide ethanol solution (KOH w
EtOH) was added. Sample tubes was kept for 2 h in a 70 °C temperature-controlled bath. Suberic acid extract was separated from bark residues by vacuum filtration. Extract-free bark residue was washed with water and oven dried at 105 °C for 24 h for later lignin analysis. For suberic acid GC-MS analysis, samples were prepared by pipetting 0.3 mL of extract into a 15 mL glass test tube and diluting the extract with 3 mL of water. Two drops of bromocresol green was added as a pH indicator and extracts acidified (pH < 3.8) with 0.25 M sulphuric acid (H2
). Two mL of internal standard C21:0/betunilol 0.2 mg/mL in methyl tert
-butyl ether (MTBE) was added and thoroughly shaken. Suberic acids were separated from extracts with liquid-liquid separation by adding again 2 mL of MTBE, vortexing thoroughly and letting phases to settle, collecting MTBE phase into a new 15 mL glass test tube and repeating the liquid-liquid extraction two more times with 3 mL of MTBE. Finally, the collected suberic acid MTBE extract was washed with 2 mL of water. Extracts were dried under a nitrogen flow in 40 °C temperature-controlled bath. Next 150 µL of silylating agent (1:4:1 mix of N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA):chlorotrimethylsilane (TMCS, Merck KGaA, Darmstadt, Germany):Pyridine) was added into test tubes and put into a 70 °C temperature-controlled bath for 45 min. Ready samples was transferred to GC-MS vials and analyzed. Suberic acid compounds were recognized by comparison with MS library values and the sum of total suberic acid content was determined by means of an internal C21:0 standard [43
Lignin content was determined from previously collected suberic acid and extract free bark samples as a sum of an acid soluble and insoluble (Klason) lignin. Duplicate oven dried samples (200 mg) were prepared by weighing them out into closable 50 mL glass test tubes and adding 2 mL of 72% sulphuric acid (H2
). After mixing thoroughly, samples were incubated in 30 °C for 1 h. After incubation, solid residues and extracts was transferred into 100 mL closable pressure and heat resistant glassware by washing samples with 56 mL of water having solid residue in 4% sulphuric acid (aq) in the end. Samples were placed in an autoclave for 1 h in 120–125 °C. After the autoclaving process, samples were allowed to cool down to ambient temperature and acid soluble and insoluble samples were separated from each other by vacuum filtration. Filtered and washed solids were dried in 105 °C for 24 h and determined gravimetrically as an acid insoluble Klason lignin. Filter collected solution containing acid soluble lignin was measured by UV-Vis spectroscopy at 240 nm wavelength and lignin content was calculated by reference to the corresponding Laboratory Analytical Procedure [44
Cellulose content was determined by acid hydrolysis from extract free bark samples. Ten mg samples were weighed out into closable 10 mL glass test tubes as duplicates. Two cellulose standards were made in the same way. A small glass ball was inserted to every test tube to help subsequent mixing. Next 200 µL of 72% sulphuric acid (H2
) was added and allowed to react for 2 h. Then 0.5 mL of water was mixed into the samples and left to react for 4 h. Finally, 6 mL of water was mixed into the samples that were left to stand at room temperature overnight. The next day, test tubes were put into an autoclave for 1 h at 120–125 °C. After the sample test tubes had cooled down to ambient temperature, 2 drops of bromocresol green pH indicator was added. Small amounts of barium carbonate were added carefully, and the mixture was vortexed thoroughly after every addition until a blue color appeared. One mL of D-sorbitol internal standard (250 mg/50 mL H2
O) was added and after mixing the samples were centrifuged. Supernatants were collected and dried under a nitrogen flow in 40 °C. After the samples were dried, the residual solvent was removed with the help of a 40 °C vacuum oven. Samples were silylated by adding in order 150 µL pyridine, 150 µL of 1,1,1,3,3,3-hexamethyldisilazane (HMDS, Sigma-Aldrich Chemie GmbH, Steinheim, Germany) and finally 70 µL of trimethylsilyl chloride (TMCS, Merck KGaA, Darmstadt, Germany). After careful mixing, samples were left to react overnight at room temperature. The next day, the clear liquid phase was collected and analyzed by GC-FID [43
Hemicellulose content and structural sugars were determined with acid methanolysis. Analyses were done as duplicate by weighing out 8–12 mg of extract free bark samples into pear shaped flasks. Standard samples were made from 1.0 mL of monosaccharide solvent mix containing arabinose, glucose, glucuronic acid, galactose, galacturonic acid, mannose, rhamnose and xylose (1.0 mg/mL each). Two mL of methanolysis reagent 2M HCl (anhydrous MeOH) was added and vortexed thoroughly before placed into oven for 5 h under 100 °C. After vials cooled down, samples were neutralized by adding 80 µL of pyridine. One mL of inner standard sorbitol 0.1 mg/mL + resorcinol 0.1 mg/mL (MeOH) was added in each sample and vortexed. After solutions settled down, 1 mL of clear phase was collected and dried out under nitrogen flow in 40 °C temperature-controlled bath. Dry sample residues were silylated by adding 150 µL pyridine, 150 µL HMDS disilazane and 70 µL TMCS in order. Samples were left overnight in room temperature and settled sample solutions clear phase was carefully collected and analyzed by GC-FID [45
4.4. Condensed Tannins
Condensed tannins (CT, proanthocyanidins) were determined by HPLC after thiolytic degradation. Samples were weighed (20–30 mg) into 1.5 mL Eppendorf vials and 1 mL of depolymerization reagent (3 g cysteamine dissolved in 56 mL of methanol acidified with 4 mL of 13 M HCl) was added. The vials were sealed, vigorously mixed (vortexer) and incubated for 60 min at 65 °C. During the incubation the samples were vortexed few seconds every 15 min. After 60 min the samples were transferred into an ice bath to stop the reaction. Samples were filtrated into HPLC vials and analyzed on an Agilent 1290 Infinity UHPLC device equipped with a Zorbax Eclipse Plus C18 column (Agilent Technologies, Santa Clara, CA, USA, 50 × 2.1 mm i.d., 1.8 μm). The binary mobile phase consisted of 0.5% formic acid (aq) and acetonitrile. Elution was started with 2% acetonitrile, isocratically for 2 min, followed by a linear gradient to 5% in 3 min, to 15% in 7 min, to 20% in 3 min, to 35% in 5 min, to 90% in 1 min, and back to the starting point in 2 min. The post-time was 2 min before the next injection. The flow rate was 0.5 mL/min and injection volume 2 μL. Elution was monitored by diode array detection (DAD; λ1 = 270 nm, λ2 = 280 nm) and fluorescence detection (FLD; λex = 275 nm, λem = 324 nm). CT degradation products, i.e., free flavan-3-ols (terminal units) and their cysteaminyl derivatives (extension units), were quantified using external standards of catechin, epicatechin, gallocatechin, epigallocatechin, and thiolyzed procyanidin B2. Cysteamine, catechin, epicatechin, gallocatechin, and epigallocatechin were purchased from Sigma-Aldrich Finland Oy (Espoo, Finland). Procyanidin B2 was obtained from Extrasynthese (Lyon, France).
4.5. Antioxidative Properties and Total Phenol Content
Samples of dried aspen bark were extracted (1:10) for the bioactivity testing with hot distilled water or 70% ethanol. Bark powder was weighed in 15 mL test tubes and boiling H2O (or 70% ethanol (aq)) was added to the tube, then vortexed for 2 min, incubated for 15 min at RT after which vortexed for 2 min, and centrifuged for 10 min at 8400 rpm (Sigma 2-16KL). After centrifugation the supernatant was removed to a clean tube and kept at +4 °C until tested.
To screen the antioxidative properties of the aspen bark fractions, four assays to cover different antioxidant mechanisms were used. The measurements were performed on 96-microplate format with a multimode microplate reader (Varioskan Flash, Thermo Scientific, Waltham, MA, USA). Detailed descriptions of the testing protocols of ORAC, FRAP and SCAV are described in the paper of Vaario et al. [46
4.5.1. Ferric Reducing Antioxidant Power (FRAP)
Single electron transfer (SET)-based FRAP (ferric reducing antioxidant power) assay measures the ability of an antioxidant to reduce ferric (FeIII) to ferrous (FeII) ion FeSO4·7H2O was used as a standard and four different dilutions of the samples were used to fit them to the standard curve.
The samples were mixed with 20 mM FeCl3
O and 10 mM 2,4,6-tris(2-pyridyl)-s-triazine (TPTZ) in 300 mM acetate buffer pH 3.6. The assay was run with three technical replicates and the absorbance was measured at 594 nm after the formation of ferrous-tripyridyltriazine complex in the reaction mixture. Reagents: FeSO4
O (Sigma-Aldrich Chemie GmbH), l
(+)-ascorbic acid (VWR Chemicals, Radnor, PA, USA), 2,4,6-tris(2-pyridyl)-s-triazine (TPTZ, Sigma-Aldrich Chemie GmbH, St. Louis, MO, USA) [47
4.5.2. Oxygen Radical Absorbance Capacity
Hydrogen atom transfer-based (HAT) the Oxygen Radical Absorbance Capacity (ORAC) assay measures the oxidative dissociation of fluorescein at the presence of peroxyl radicals, which causes reduction in the fluorescence signal. The antioxidative ability of an extract or a compound is measured as the inhibition of the fluorescein breakdown. The method used here was modified to 96-well format from those described by Huang et al. [48
] and Prior et al. [49
] with two technical replicates of each sample on the plate.
The sample in 0.075 M phosphate buffer pH 7.5 was mixed with 8.16 × 10−5 mM fluorescein and 2,2′-azobis(2-methylpropionamidine) dihydrochloride). A series of twenty dilutions was used for each sample to adjust the sample concentration to the standard curve. Trolox was used as a standard and the results are expressed as Trolox ((±)-6-Hydroxy-2,5,7,8,-tetramethylchromane-2-carboxylic acid, a vitamin E analog) equivalents (µmol/L TE). Reagents: Phosphate buffer pH 7.5 (Merck), fluorescein (Sigma-Aldrich Chemie GmbH, St. Louis, MO, USA), 2,2′-azobis(2-methylpropionamidine) dihydrochloride (Sigma-Aldrich Chemie GmbH, St. Louis, MO, USA). Trolox ((±)-6-hydroxy-2,5,7,8,-tetramethylchromane-2-carboxylic acid, vitamin E analog) (Sigma-Aldrich Chemie GmbH, St. Louis, MO, USA).
4.5.3. H2O2 Scavenging Assay
scavenging assay, based on transition metal chelation, was modified from Hazra et al. [50
] and Jiang et al. [51
] with microplate reader in 96-well format with four technical replicates on each plate. In the assay the formation of ferric-xylenol orange complex indicates the ability of the sample to scavenge H2
and prevent the oxidation of Fe(II) to Fe(III). Each sample was measured using the extract (1:10) without dilutions and the result is expressed as inhibition percentage (%) of Fe(II) oxidation to Fe(III).
In the assay the sample is mixed with an aliquot of 2 mM H2O2, 2.56 mM ammonium iron (II) sulphate·6H2O and 111 µM xylenol orange disodium salt and after 30 min incubation, the absorbance of ferric-xylenol orange complex at 560 nm was measured.) Sodium pyruvate (Sigma-Aldrich Chemie GmbH, St. Louis, MO, USA) was used as a reference compound. Reagents: H2O2 (Merck KGaA, Darmstadt, Germany), ammonium iron (II) sulphate·6H2O (BDH Prolabo, Dubai, UAE), xylenol orange disodium salt (Sigma-Aldrich Chemie GmbH, St. Louis, MO, USA), sodium pyruvate (Sigma-Aldrich Chemie GmbH, St. Louis, MO, USA).
4.5.4. Folin-Ciocalteu Assay
Folin-Ciocalteu assay [52
] was used to analyze the total phenolic content in the extracts. The samples were mixed with Folin-Ciocalteu reagent and Na2
and absorbance measured at 750 nm with gallic acid as a reference compound. The results are expressed as gallic acid equivalents per 1 g (mg GAE/g). Reagents: Na2
(Merck KGaA), gallic acid (Sigma-Aldrich Chemie GmbH, St. Louis, MO, USA), ethanol (Altia Industrial, Helsinki, Finland), Folin-Ciocalteu reagents (Merck KGaA, Darmstadt, Germany).
4.6. Ash Content
Ash content was determined with a modified TAPPI protocol [55
]. One g of fine-grained samples was weight out and dried in 105 °C for 24 h. Bone dry powder was weighed and placed into the furnace. The ashing furnace temperature program was set to rise 80 °C per hour from ambient to 270 °C, then 220 °C per hour to 470 °C and finally 100 °C per hour to the target temperature of 540 °C and kept for 180 min. After the furnace had cooled down, the remaining incombustibles were determined gravimetrically and calculated as ash content.