- Email: [email protected]
Bottom ash of trees from Cameroon as fertilizer
Applied Geochemistry 72 (2016) 88e96
Contents lists available at ScienceDirect
Applied Geochemistry journal homepage: www.elsevier.com/locate/apgeochem
Bottom ash of trees from Cameroon as fertilizer €lle Trouve c, Reto Giere d Christoph Maschowski a, *, Marie Claudine Zangna b, Gwenae €t, 79104 Freiburg, Germany Institut für Geo und Umweltnaturwissenschaften, Albert-Ludwigs-Universita ACK e Alumni Club Kamerun, Yaound e, Cameroon c Laboratoire Gestion des Risques et Environnement, Universit e de Haute-Alsace, 68093 Mulhouse Cedex, France d Department of Earth and Environmental Science, University of Pennsylvania, Philadelphia, PA 19104-6316, USA a
b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 27 April 2016 Received in revised form 4 July 2016 Accepted 6 July 2016 Available online 9 July 2016
Utilization of wood bottom ash as fertilizer additive contributes to the return of valuable nutrients to agricultural soils, especially when no artificial mineral fertilizer is being used. In general, wood combustion ash is enriched in calcium and potash, and may also contain elevated amounts of zinc, but the concentrations of these elements depend on tree species, part of the tree, harvest season and local soil type. In this study, bottom ash samples from eight different agricultural wood species from Cameroon, Africa were investigated by using X-ray diffraction and atomic absorption spectroscopy to determine the refractory components and the concentrations of selected heavy metals and arsenic. Results show calcite, potassium salts, periclase and quartz as major components. These phase contents were used to calculate major element concentrations, which were subsequently validated by X-ray-fluorescence analysis. The chemical compositions varied within the range of common compositions of wood ashes. Six of the ashes reached sufficient concentrations of calcium to be defined as a “calcium fertilizer”. Pb contents are most variable, ranging from 0.03 to 21.1 mg/kg. Concentrations of Ni, Cu, Zn, Cd, Pb, and As are all lower than the strictest limit concentrations required for wood ash fertilizers and therefore, the studied wood ashes can be used without environmental concern. © 2016 Elsevier Ltd. All rights reserved.
Keywords: Biomass Wood ash Fertilizer Soil amendment
1. Introduction Solid biomass, e.g. wood, contains various inorganic components in different amounts and speciation, depending on plant type, plant part, soil, and harvest season (Vassilev et al., 2010). Inorganic constituents remain as ash after combustion, as they are non-combustible. Depending on the level of completeness of the combustion, the ash may contain some organic compounds (Vassilev and Vassileva, 1997). The ash is separated into two fractions: one remains on or below the combustion grate as bottom ash; the other is transported along with the flue gas as fly ash, and it comprises the minor part in open fireplaces. Bottom ashes are considered to be unproblematic and sometimes are used as fertil€ ser, izer additives, because they contain valuable nutrients (Ro 2008). Fly ashes, however, may be enriched in (toxic) heavy metals and metalloids and therefore, have to be disposed as problematic waste, after being trapped by a flue gas treatment device. The suitability of bottom ashes for utilization as fertilizer
* Corresponding author. E-mail address: [email protected] (C. Maschowski). http://dx.doi.org/10.1016/j.apgeochem.2016.07.002 0883-2927/© 2016 Elsevier Ltd. All rights reserved.
additive/soil amendment strongly depends on fuel type and soil quality, and has to be assessed for each species and locality (Vassilev et al., 2013a,b). Therefore, the characterization of a wide range of biomass ash types can help in understanding the nutrient cycle and trace element uptake processes, especially when it is crucial to return ashes from biomass combustion to the soil in order to avoid depletion of nutrients. Within the living memory, potash is known as vital nutrient for plant growth, as it consists essentially of water-soluble potassiumbearing salts, which are completely bioavailable (Sharifi et al., 2013). Because the potassium content in wood ash can be as high as 50% by weight (calculated as K2O), this type of ash is being used as a source of potassium for the production of fertilizers, glass (Tite et al., 2006) and soap (Farmer, 2013). However, in modern agriculture, ash from wood combustion is no longer considered a feasible source of potassium, but wood ash might be the only potassium source, when no artificial/mineral fertilizer is used and the soil type is depleted in this element, e.g. when no clay-minerals are present (Barre et al., 2007). Ash from wood combustion is depleted in nitrogen and there€ser, 2008), but can be fore, cannot be used as a general fertilizer (Ro
C. Maschowski et al. / Applied Geochemistry 72 (2016) 88e96
taken as a special amendment to deliver alkaline earth metals, which are able to neutralize acidification of the soil caused by acid rain, i.e. atmospheric pollution (Clarholm, 1994). In general, the calcium concentrations in wood ash are relatively high (Ca mostly present in the form of calcium carbonate and/or calcium oxide), and therefore, wood ash may be used as liming agent to neutralize acid soils and increase the activity of soil bacteria (Bache and Sharp, 1976). In addition, some wood ash types contain enough sulfur and phosphorous that they could be considered as a source of these macro-nutrients. In this study, samples of wood bottom ash from eight different agricultural tree types from Cameroon were investigated by using X-ray diffraction (XRD) with subsequent Rietveld refinement, and by X-ray fluorescence (XRF) and atomic absorption spectroscopy (AAS). The main goals of this study were to identify the crystalline phases and their relative proportions, to evaluate the suitability of combining different analytical techniques, and to compare the concentrations of some troublesome heavy metals and arsenic against strictest limit values for wood ash and other mineral fertilizers (Anon., 2002) (Vesterinen, 2003) in order to find out whether or not the ashes could be used as fertilizers in Cameroon. 2. Samples and methods All trees were harvested in January 2011 in the bush-region of the village Evondo near the international airport ‘Nsimalen’ of , Cameroon (geo-coordinates: 3.743300, 11.524313). To Yaounde represent a real-life scenario, the wood specimens consisted of a mixture of different tree parts (stem wood with bark and twigs). All tree samples were stored for one year prior to combustion for drying reasons. The soil in this region is classified as a “Ferralsol” according to the Harmonized World Soil Database (IIASA, 2012), which stands for iron- and aluminum-rich soil types, often depleted in nutrients. The geological basement underlying the soil consists of metamorphic rocks, represented by embrechite (metamorphic migmatite) and gneiss (Geological map of Cameroon (BMArchives, 2015)). The ash samples studied here were produced by combustion in open fireplaces of eight tree species (Fig. 1), which were burned in , Cameroon. The ashes represent bottom ash the vicinity of Yaounde and were collected directly after combustion, following cooling. All ash samples were ground in an agate mortar and analyzed by XRD with a Cu-Ka X-ray source, using a high-precision setting (2e75 2q; step size of 0.005 ; 5s/step) on a BRUKER D8 Advance diffractometer at the University of Freiburg, Germany. The raw data were processed with pattern matching software (BRUKER DIFFRAC.EVA) in order to identify the phases present (Fig. 2). For each of the identified phases, a structure file was obtained from the American Mineralogist Crystal Structure Database (Downs and Hall-Wallace, 2003). These structure files were then applied to the datasets in the Rietveld refinement program DIFFRACsuite TOPAS by Bruker ltd. to determine the relative proportions of crystalline phases present (Fig. 3). The amorphous content was determined from background signal (Scarlett and Madsen, 2006). Quality of the data retrieved from XRD and Rietveld refinement depends on various aspects: sample preparation (grain-size distribution, surface roughness), degree of crystallinity (amorphous content, e.g. glassy substances and organic matter), and strategy of the refinement procedure (e.g. (Ward and French, 2006)). Grain-size of the sample depends on the preparation method. In our case, we ground it manually with the help of an agate mortar, which is a working practice, but not as accurate as automated grinding. Surface roughness depends on the technique of distributing the sample into the mold of the sample holder, which was done manually as well.
89
Quantification of the amorphous content is usually achieved by spiking the sample with a specific concentration of a known substance, which is 100 percent crystalline. From this information, Rietveld refinement software tools can easily define the degree of crystallinity of the sample. Because of limited sample quantity, we decided to perform XRD measurements without spiking the sample, so that enough uncontaminated material was kept for subsequent XRF and AAS analyses. The calculation of the amorphous content, however, is important; in our case it was done by adding an artificial one-peak phase representing the background signal and thus the amorphous part of the sample. The uncertainty of this method cannot be determined, but the applied method seems to work quite well for the ashes of this study. The refinement procedures were performed without comprehensive corrections like the compensation of preferred orientation, lattice parameters and micro-strain. The bulk chemical composition of the ash samples was analyzed by SGS Lakefield, a Canadian laboratory with ISO17025 standard specializing in geochemical analysis. The specific XRF-analysis method used for this project was the BORATE FUSION/XRF WHOLE ROCK PACKAGE (GO XRF76V). To obtain accurate data for selected heavy metals and arsenic, the samples were also analyzed by AAS at the University of Freiburg, Germany, using an AAS Vario 6 instrument with graphite tube by Analytik Jena. The samples were first digested with a mixture of nitric acid and hydrochloric acid (aqua regia) under microwave treatment. 3. Results The macroscopic colors of the ashes are relatively light for all tree types (see photographs in Fig. 1). The color descriptions given in Fig. 1 conform to the RAL color matching system used in Europe. The light color reflects the presence of only small amounts of organic matter residue, i.e. the combustion process was quite complete. Therefore, the carbon content was not determined and will not be discussed further in this study. 3.1. Mineralogical composition All ash samples contain a considerable amount of calcite (CaCO3), ranging from 25.8 to 70.7 wt% (percent by weight ¼ g $ 100 g1) (Table 1, Fig. 4). They are further characterized by relatively low contents of arcanite (K2SO4), which ranges from 4.17 wt% (mango tree) to 9.26 wt% (cheesewood). Other potassium phases identified include a chloride (sylvite) and a K-bearing carbonate (fairchildite), which is the most variable component, ranging from undetectable amounts (abachi tree) to 38.8 wt% (African drupe tree). The concentration of free lime (CaO) is unexpectedly low (4.42 wt%); in the ashes produced through combustion of avocado tree and African tulip tree the phase was not even detected by XRD (Table 1). Periclase (MgO) is the only magnesium phase present, and its concentration may be as high as 13.1 wt% (African tulip tree). Similarly, halite (NaCl) is the only sodium-bearing phase, but its concentration is very low, except for the ash derived from the abachi tree (12.3 wt%). The content of quartz (SiO2) is highly variable among the different samples (1.67e21.7 wt%), as is the content of amorphous material (7.49e41.4 wt%). 3.2. Bulk chemical composition 3.2.1. Oxide concentrations derived from XRD data In order to assess the utilization potential and quality of the ashes as fertilizer additives, the concentrations of all phases
90
C. Maschowski et al. / Applied Geochemistry 72 (2016) 88e96
Fig. 1. Names and photographs of the source tree species and corresponding bottom ashes analyzed in this study.
containing either Na, or K, or Mg, or Ca, or S have been taken to calculate the corresponding bulk-oxide concentrations. Calculated concentrations of, Na2O, K2O, MgO, CaO, and SO3 were then compared to the minimum concentrations that a potential fertilizer material has to reach. As reference values, we chose as an example the minimum concentrations required for utilization of other r and Haasnoot, ‘inorganic fertilizers’ in the Netherlands (Sarabe 2012), because they also conform to the generally complex regulations for fertilizers in other countries, such as, Germany (Düngegesetz DüG) and Switzerland (Dünger-Verordnung, DüV). Because of the relatively high content of amorphous constituents in some samples and because XRD and Rietveld refinement can only determine crystalline phases, all calculated oxide concentrations
(Table 2) are minimum values. Six of the eight studied samples contained CaO in calculated concentrations that exceed the minimum concentration limit of r and Haasnoot (2012). The two 25.0 wt%, as suggested by Sarabe exceptions, avocado tree and cheesewood ashes, however, contain significantly higher amounts of amorphous components than the other samples (Table 1). This results in underestimation of oxide contents calculated from the XRD data. The calculated contents of the other oxides are below the minimum values suitable for fertilizer additives for all samples. The Na2O concentrations are very low (~1 wt%) or not detectable, except for the Abachi tree ash (6.50 wt%). Although the calculated K2O concentrations are lower than 25 wt%, they are highly variable and range from 2.49 wt% for
C. Maschowski et al. / Applied Geochemistry 72 (2016) 88e96
91
Fig. 2. XRD patterns with phase identification of mango tree ash and umbrella tree ash. Halite and sylvite were not detected in these two ashes and therefore were not labeled.
abachi tree ash to 19.4 wt% for African drupe tree ash. The SO3 content is low, whereas the MgO content is variable and is highest in African tulip tree ash (13.1 wt%). 3.2.2. X-ray fluorescence (XRF) data To evaluate the bulk-oxide concentrations calculated from the XRD and Rietveld refinement data, the ashes were also analyzed by XRF. The values for loss on ignition (LOI) were unexpectedly high, up to 36 wt% (Table 3). Depending on the tree type, the SiO2 content in the ashes is between 7.37 wt% (African tulip tree) and 29.7 wt% (mango tree), which results in the highest variance compared with all other components (Table 3). In contrast, Al2O3 and Fe2O3 contents are low and fairly uniform in the ashes of all tree types, with maximum concentrations observed for cheesewood ash (Al2O3: 4.39 wt%; Fe2O3: 3.06 wt%). MgO concentrations range from 2.94 wt% (mango tree) to 8.42 wt% (African tulip tree). Contents of CaO are high (up to 39.1 wt% for umbrella tree) and, except for the avocado tree and cheesewood ashes, exceed the minimum concentration for utilization as a calcium fertilizer. Na2O content is below 0.2 wt% for all samples. K2O contents, however, are considerably higher and quite variable, ranging from 5.22 wt% (abachi tree) to 15.6 wt% (cheesewood). P2O5 contents are relatively low, ranging from 0.51 wt% for abachi tree to 2.62 wt% for African tulip tree. Comparison of the XRF data (Table 3) with the data derived from XRD and Rietveld refinement (Table 2) reveals that absolute differences in concentration are mostly less than ±5 wt% (Table 4). The
relative differences, however, are high for some of the ash samples, especially in terms of the Na2O content (e.g. avocado, African tulip and abachi trees), because the absolute concentrations of Na2O are so low (Table 3). Because of these relatively low differences, which are due to analytical deviations (semi-quantitative XRD-results are associated with a relative uncertainty of approximately five percent), the probable compositions of the amorphous phases cannot be predicted. Bulk chemical compositions of the eight tree ash samples were compared to those of ashes from the standard DIN þ wood pellet (DINþ is a combination of the German DIN 51731 standard and the € Austrian ONORM M 7135 standard), combusted in a small-scale private wood pellet boiler (50 kW nominal thermal output), and from common wood chips, combusted in a medium-scale biomass power plant (1.7 MW nominal thermal output) located at St. Peter in the Black Forest, Germany. The wood chips were a mixture of softwood species (mainly spruce) from merchantable wood consisting of branches with a diameter >70 mm. Composition of the wood chip ash lies within the compositional range of the eight tree ash samples studied, whereas the pellet ash is depleted in all majorelement oxides, except for SiO2, whose concentration is about three times higher (Fig. 5). The results from XRF analysis provide similar information about the utilization potential of the ashes as fertilizer additives: all ash samples satisfy the requirement for calcium contents, except for the avocado tree ash and cheesewood ash. This result is consistent with that obtained from the XRD and Rietveld analysis.
92
C. Maschowski et al. / Applied Geochemistry 72 (2016) 88e96
Fig. 3. XRD patterns of mango tree ash and umbrella tree ash with Rietveld refinement applied.
Table 1 Semi-quantitative mineralogical composition (in wt%), as determined by XRD and subsequent Rietveld refinement of the eight tree ashes. Mineral species
Mango tree
African drupe tree
Avocado tree
African tulip tree
Dove-wood
Abachi tree
Umbrella tree
Cheese-wood
Lime (CaO) Calcite (CaCO3) Arcanite (K2SO4) Quartz (SiO2) Fairchildite (K2Ca(CO3)2) Periclase (MgO) Sylvite (KCl) Halite (NaCl) Amorphous
4.42 52.3 4.17 12.9 1.43 3.39 n.d. n.d. 21.3
1.67 31.6 7.14 1.67 38.8 5.63 n.d. n.d. 13.4
n.d. 28.0 7.06 21.7 2.85 n.d. n.d. 1.92 38.4
n.d. 47.6 8.63 6.89 2.55 13.1 1.74 1.09 18.5
1.66 56.3 5.87 6.96 1.14 6.64 7.91 n.d. 13.5
2.50 60.6 4.57 5.33 n.d. 7.25 n.d. 12.3 7.49
1.12 70.7 6.25 3.73 1.59 5.20 n.d. n.d. 11.4
0.82 25.8 9.26 13.7 1.21 6.10 1.70 n.d. 41.4
3.3. Heavy metals and arsenic concentrations Because the contents of the heavy metals Ni, Cu, Zn, Cd, and Pb, and of As in the studied ash samples were below the XRF detection limits, these elements were analyzed by AAS. Data (Table 5) reveal considerable ranges in concentration for Cu, Zn, and Pb. Cu content is as low as 64.6 mg/kg in umbrella tree ash and as high as 230 mg/ kg in dovewood ash. Zn ranges from 58.1 mg/kg in African drupe tree ash to 810 mg/kg in avocado tree ash, which nearly suggests this tree species to be a hyper-accumulator for zinc (Rascio and Navari-Izzo, 2011). Pb content exhibits the greatest variability, ranging from 0.03 mg/kg in African drupe tree ash to 21.1 mg/kg in dovewood ash. There is no qualitative correlation between the concentrations of Cu, Zn, and Pb in the different tree ashes, except for dovewood ash and cheesewood ash, which exhibit relatively high abundances of all three heavy metals. Ni exhibits the smallest compositional variation, ranging from 19.3 mg/kg in umbrella tree ash to 34.9 mg/ kg in cheesewood ash. Cd is not detectable by AAS in five of the eight samples and is very low for the others, with the highest
concentration of 0.85 mg/kg in mango tree ash. As is relatively constant too, ranging from 0.79 mg/kg in African drupe tree ash to 2.31 mg/kg in abachi tree ash. As shown in Fig. 6 and Fig. 7, concentrations of the studied heavy metals and arsenic are all below the limit values, which represent the lowest values selected from the national regulations of different countries. Lowest limit values found were those from Sweden (Anon., 2002) and Finland (Vesterinen, 2003). 4. Discussion X-ray diffraction with subsequent Rietveld refinement appears to be a fast and cost-effective method to analyze ashes from biomass combustion because the crystalline refractory materials, the main components of ash, can be determined fairly easily. Furthermore, it allows for the differentiation of phases, which in turn can help in assessing burning temperature and atmosphere €ller et al., 2015). For example, the low concentration of free (Reinmo lime suggests that the combustion temperature was at the very low end of the temperature range for calcite decomposition, which
C. Maschowski et al. / Applied Geochemistry 72 (2016) 88e96
93
Fig. 4. Relative compositions by weight of mineral species in the studied ash samples, as determined by XRD and subsequent Rietveld refinement.
Table 2 Comparison of selected oxide concentrations (in wt%) in the ash samples (calculated from the contents of crystalline phases obtained from XRD and Rietveld refinement) with r and Haasnoot, 2012)); bold values > Cmin fert.; n.d. ¼ not the minimum concentration limits for utilization as ‘other inorganic fertilizers’, labeled as Cmin fert. (data from (Sarabe detected.
Na2O K2 O MgO CaO SO3
Cmin fert.
Mango tree
African drupe tree
Avocado tree
African tulip tree
Dove-wood
Abachi tree
Umbrella tree
Cheese-wood
50 25 15 25 25
n.d. 2.85 3.39 34.1 1.90
n.d. 19.4 5.63 28.5 3.25
1.02 4.99 n.d. 16.4 3.21
0.58 7.95 13.1 27.2 3.93
n.d. 13.7 6.64 33.5 2.67
6.50 2.49 7.25 36.5 2.08
n.d. 4.04 5.20 41.1 2.84
n.d. 7.69 6.10 15.6 4.21
Table 3 Bulk chemical composition (in wt%) of the studied ash samples, as determined by XRF analysis with the minimum concentration limits for utilization as ‘other inorganic r and Haasnoot, 2012); bold values > Cmin fert. fertilizers’, labeled as Cmin fert. (data from Sarabe Cmin fert. SiO2 Al2O3 Fe2O3 MgO CaO Na2O K2 O P2O5 MnO LOI SUM
15 25 50 25
Mango tree
African drupe tree
Avocado tree
African tulip tree
Dove-wood
Abachi tree
Umbrella tree
Cheese-wood
Vari-ance
29.7 2.75 1.46 2.94 31.0 0.10 5.52 0.81 0.44 23.8 98.8
27.5 1.17 0.83 5.04 26.0 0.04 11.0 1.50 0.06 21.2 94.5
23.5 3.65 2.38 5.57 17.5 0.19 13.9 1.68 0.29 29.5 98.4
7.37 1.33 1.28 8.42 30.3 0.07 13.9 2.62 0.03 31.4 97.0
16.6 2.14 2.39 4.81 31.9 0.16 7.43 2.36 0.68 28.6 97.2
14.0 2.20 1.93 5.24 36.7 0.08 5.22 0.51 0.13 29.4 95.6
7.89 1.07 0.85 4.06 39.1 0.06 8.28 0.83 0.18 36.0 98.4
20.4 4.39 3.06 4.05 18.6 0.18 15.6 1.87 0.09 29.8 98.5
70.8 1.45 0.64 2.59 60.7 <0.04 16.5 0.58 0.05 20.5 2.42
Table 4 Comparison of results from XRD and XRF analyses for selected elements as oxides with values showing the differences in wt%; positive values indicate overestimation and negative values indicate underestimation of concentrations from XRD calculations.
Na2O K2 O MgO CaO
Mango tree
African drupe tree
Avocado tree
African tulip tree
Dove-wood
Abachi tree
Umbrella tree
Cheese-wood
0.10 2.67 0.45 3.08
0.04 8.42 0.59 2.47
0.83 8.91 5.57 1.13
0.51 5.95 4.64 3.06
0.16 6.30 1.83 1.57
6.42 2.73 2.01 0.25
0.06 4.24 1.14 2.02
0.18 7.91 2.05 3.03
takes place between 600 and 850 C (Grapes, 2010), (RodriguezNavarro et al., 2009). This conclusion is supported by the relatively high abundance of calcite in all samples (see Table 1). Another contributing factor to the presence of calcite could be the
reaction of CO2 from ambient air with CaO to form secondary calcite. Calcium, however, is also present in another carbonate, fairchildite. These observations suggest that the leaching behavior or liming potential of the studied ashes is similar to any other
94
C. Maschowski et al. / Applied Geochemistry 72 (2016) 88e96
Fig. 5. Comparison of the composition of the ash samples and other standard biomass ash samples.
Table 5 AAS results showing bulk concentrations (in mg/kg) of selected heavy metals and arsenic in the studied ashes; safety values taken from limit values of Ni, Cu, and As in wood ash as fertilizer from Sweden, Zn, Cd, and Pb from Finland.
Ni Cu Zn Cd Pb As
Safety limits
Mango tree
African drupe tree
Avocado tree
African tulip tree
Dove-wood
Abachi tree
Umbrella tree
Cheese-wood
Detection limit
70 400 1500 3 150 30
22.9 85.3 325 0.85 0.11 1.49
23.3 110 58.1 0.27 0.03 0.79
23.9 136 810 0.48 0.13 1.35
24.2 92.6 70.5 <0.02 7.74 1.26
28.9 230 495 <0.02 21.1 1.55
21.5 126 71.81 <0.02 7.86 2.31
19.3 64.6 136 <0.02 4.93 1.60
34.9 159 328 <0.02 11.3 1.25
0.0376 0.0141 0.0080 0.0046 0.0009 0.0012
Fig. 6. Concentrations of selected heavy metals (Ni, Cu, Zn) in the studied ash samples. All concentrations are lower than the lowest limit values which could be found in the literature (limit values of Ni and Cu in wood ash as fertilizer from Sweden, Zn from Finland).
calcium fertilizer. Concentrations of the potassium-bearing phases arcanite and fairchildite, hence the potash content, are relatively low in all studied ashes, except for the ash from the African drupe tree (Table 1). Therefore, the potassium content is of lesser importance in terms of nutrient supply and has a minor, non-deciding influence on the evaluation of the utilization potential of the ashes. In addition to decomposition and dissociation of mineral species during combustion, some elements (e.g. K, S, Na, Cu, Zn, Pb, and As) tend to become volatile as a function of increasing temperature, whereas others (e.g. Mg, P, Mn, Al, Fe, and Si) are more refractory and tend to accumulate in the bottom ash during combustion at temperatures below 1400 C (Misra et al., 1993). Because Zn is an
important macro-nutrient, its content has to be considered. According to Verbruggen et al. (2009), plants with zinc concentrations >10 mg $ g1 (1 wt%, dry basis) qualify as hyper-accumulators. From our bottom ash data, however, we cannot deduce the absolute Zn concentrations in the plants prior to combustion. As the bottom ash from the avocado tree contains such high levels of zinc, we conclude that the avocado tree might be a hyper-accumulator plant. In this study, no correlation of zinc with the concentrations of other elements could be observed. When the aim is that zinc should act as a nutrient, it is important to achieve a pH lower than 7.3 in the soil to avoid the blocking of zinc uptake by plants (Meyer, 1999). The relatively low Na2O contents in all ashes can be attributed to the volatile character of Na. This result is favorable when
C. Maschowski et al. / Applied Geochemistry 72 (2016) 88e96
95
Fig. 7. Concentrations of selected heavy metals (Cd, Pb) and As in the studied ash samples. All concentrations are lower than the lowest limit values found in the literature (limit values of Cd and Pb in wood ash as fertilizer from Finland, As from Sweden). The limit value of Pb is not shown in this figure, because of graphical reasons e it is one order of magnitude higher (150 mg/kg, Finland) than the Pb content of the ash samples.
considering the fertilizer potential of the ashes, because Na is not wanted in soils in significant amounts as it lowers pH, has negative effects on the growth of Na-sensitive plants and can lead to nutrient deficiencies (Davis et al., 2007). The low SO3 content in the ashes can also be attributed to its volatile character at the temperatures of combustion. It is presumed that most of the initial sulfur in the wood was transformed into potassium sulfate (arcanite) during combustion. These potassium sulfate crystals have probably been enriched in the fly ash fraction. Alternatively, they might be present as micro-crystals in the bottom ash, where they cannot be detected by XRD and thus are ‘hidden’ in the amorphous fraction of the ashes. Aluminum is most likely hosted by an amorphous phase, because no crystalline, Al-bearing compounds (e.g. clay minerals) were detected by XRD. The origin of aluminum in wood ash is often attributed to extraneous grains of clay minerals from soil, but the temperature of thermal decomposition of clay minerals (e.g. 800 C for illite (Araújo et al., 2004)) was not reached in our case, as indicated by the abundance of calcite. Because no clay mineral peaks could be found and the decomposition temperature of clay minerals was not reached, we can conclude that no clay minerals were present in any of the ash samples. The amorphous fraction primarily contains glassy substances, including Al- and Si-bearing compounds, but most likely also secondary micro-crystalline phases, including salts (possibly arcanite), as indicated by scanning electron microscope investigations on other wood ash samples (Maschowski, in prep.). The relatively high level of amorphous material in the cheesewood ash may lead to underestimation of the calcium content in calculations from XRD data. Therefore, it is concluded, that a notable amount of calcium may be located in the amorphous part for this sample. High LOI values are most likely due to high water contents, as wood ashes are in general highly hygroscopic (Martin et al., 2013), and not the result of residual carbon, as indicated by the light color of all ash samples (Fig. 1). Furthermore, decomposition of calcite during LOI determination process may contribute to the weight loss.
The mean crystallite sizes of all phases were in an acceptable range (from 100 nm for arcanite to 500 nm for quartz, approximately) and the goodness of fit (GOF) values range from 6.65 for mango tree ash to 8.33 for umbrella tree ash (Fig. 3). These GOFvalues are not far from five, which represents an optimized refinement (Kniess et al., 2012). The concentration of troublesome trace elements, i.e. As, Pb, Cd, Ni, Zn, and Cu might be controlled by the accumulation potential of each particular tree species, as well as by input through regional atmospheric pollution, contamination during harvesting, transportation, and storage. 5. Conclusions In conclusion, we propose a general applicability in terms of sufficient nutrient supply as calcium fertilizer for the six analyzed tree species ashes with minimum of 25 wt% of CaO. The other two ashes, which do not reach this level (avocado tree ash and cheesewood ash), can be used as well, but they may not fully satisfy the needs when calcium is required, for example, to increase pH in acidic soils. Concentrations of the main elements of environmental or health concern (Ni, Cu, Zn, Cd, Pb, and As) are below the lowest limits and they add up to <1 g/kg in all ashes (Fig. 8), and therefore, we propose that the studied wood bottom ashes are suitable as fertilizer and/or fertilizer amendment to be mixed with compost or with one another for use in the local agriculture and forestry of Cameroon. Calcium speciation in the ashes, where the element occurs mainly as calcite, suggests that the ashes can be applied on moderately acid soils. Limit values of troublesome elements, such as toxic heavy metals and arsenic, are for fertilizer substances only, and therefore, the ashes have to be treated with care just like any other waste material. There are no data for properties of the ashes that might be health-relevant. Although the results depend on the tree species, the uptake and ash content of inorganic substances can vary strongly depending on the soil type of the local growing area, atmospheric pollution, contamination during harvesting, transport,
96
C. Maschowski et al. / Applied Geochemistry 72 (2016) 88e96
Fig. 8. Stacked columns diagram of absolute concentrations of selected heavy metals (Ni, Cu, Zn, Cd, Pb) and As of the ash samples.
and storage. In this study, we have shown an applicable and fast-forward procedure of assessing the reuse potential of bottom ashes obtained through uncontrolled combustion of agricultural tree types in an open fireplace. Acknowledgements This work would not have been possible without the financial support by the Europen Union through its Interreg-IV Program (Oberrhein, project C35 BIOCOMBUST; www.biocombust.eu). References Anon., 2002. Recommendations for the Extraction of Forest Fuel and Compensation € nko €ping, p. 20. Fertilising. National Board of Forestry (Skogsstyrelsen), Jo Araújo, J. H. d., Silva, N. F. d., Acchar, W., Gomes, U.U., 2004. Thermal decomposition of illite. Mater. Res. 7 (2), 359e361. IIASA, F. a, 2012. Harmonized World Soil Database (Version 1.2, FAO/IIASA/ISRIC/ ISSCAS/JRC). Rome, Italy and Laxenburg, Austria: FAO and IIASA. Bache, B.W., Sharp, G.S., 1976. Characterization of mobile aluminium in acid soils. Geoderma 15 (2), 91e101. Barre, P., Velde, B., Abbadie, L., 2007. Dynamic role of “illite-like” clay minerals in temperate soils: facts and hypotheses. Biogeochemistry 82 (1), 77e88. BMArchives, 2015. Geological Map of Cameroon [Online] Available at: [Accessed 21 6 2015]. http://www.bmarchives.org/items/show/100202367. Clarholm, M., 1994. Granulated wood ash and a ‘N-free’ fertilizer to a forest soildeffects on P availability. For. Ecol. Manag. 66 (1e3), 127e136. Davis, J.G., Waskom, R.M., Bauder, T.A., Cardon, G.E., 2007. Managing Sodic Soils. s.l.. Colorado State University Cooperative Extension. Downs, R.T., Hall-Wallace, M., 2003. The American Mineralogist crystal structure database. Am. Mineralogist 88 (1), 247e250. Farmer, R.H., 2013. Chemistry in the Utilization of Wood: Pergamon Series of Monographs on Furniture and Timber. s.l.. Elsevier. Grapes, R., 2010. Pyrometamorphism. s.l.. Springer Science & Business Media. Kniess, C.T., Lima, J.C., Prates, P.B., 2012. The Quantification of Crystalline Phases in Materials: Applications of Rietveld Method. s.l.. INTECH Open Access Publisher. Martin, M., Tritscher, T., Jur anyi, Z., Heringa, M.F., Sierau, B., Weingartner, E., vo ^t, A.S.H., Baltensperger, U., Lohmann, U., 2013. HyChirico, R., Gysel, M., Pre groscopic properties of fresh and aged wood burning particles. J. Aerosol Sci. me 56, 15e29.
Meyer, R.D., 1999. Agricultural Use of Wood Ash in California. s.l.. University of California, Agriculture and Natural Resources. Misra, M.K., Ragland, K.W., Baker, A.J., 1993. Wood ash composition as a function of furnace temperature. Biomass Bioenergy 4 (2), 103e116. Rascio, N., Navari-Izzo, F., 2011. Heavy metal hyperaccumulating plants: how and why do they do it? and what makes them so interesting? Plant Sci. 180 (2), 169e181. € ller, M., Klinger, M., Schreiner, M., Gutte, H., 2015. Relationship between ash Reinmo fusion temperatures of ashes from hard coal, brown coal, and biomass and mineral phases under different atmospheres: a combined FactSage™ computational and network theoretical approach. Fuel 151 (0), 118e123. Rodriguez-Navarro, C., Encarnacion, R.A., Ana, L., Rodriguez-Navarro, A.B., OrtegaHuertas, M., 2009. Thermal decomposition of calcite: mechanisms of formation and textural evolution of CaO nanocrystals. Am. Mineral. 94 (4), 578e593. €ser, D., 2008. Sustainable Use of Forest Biomass for Energy: a Synthesis with Ro Focus on the Baltic and Nordic Region. s.l.. Springer Science & Business Media. r, A., Haasnoot, K., 2012. Recycling of biomass ashes in the Netherlands. Sarabe Proceeding of Conference ASH 2012. Scarlett, N.V.Y., Madsen, I.C., 2006. Quantification of phases with partial or no known crystal structures. Powder Diffr. 21 (04), 278e284. Sharifi, M., et al., 2013. Evaluation of liming properties and potassium bioavailability of three Atlantic Canada wood ash sources. Can. J. Plant Sci. 93 (6), 1209e1216. Tite, M.S., Shortland, A., Maniatis, Y., Kavoussanaki, D., Harris, S.A., 2006. The composition of the soda-rich and mixed alkali plant ashes used in the production of glass. J. Archaeol. Sci. 33 (9), 1284e1292. Vassilev, S.V., Vassileva, C.G., 1997. Geochemistry of coals, coal ashes and combustion wastes from coal-fired power stations. Fuel Process. Technol. 51 (1), 19e45. Vassilev, S.V., Baxter, D., Andersen, L.K., Vassileva, C.G., 2010. An overview of the chemical composition of biomass. Fuel 89 (5), 913e933. Vassilev, S.V., Baxter, David, Andersen, Lars K., Vassileva, Christina G., 2013a. An overview of the composition and application of biomass ash. Part 1. Phaseemineral and chemical composition and classification. Fuel 105 (0), 40e76. Vassilev, S.V., Baxter, David, Andersen, Lars K., Vassileva, Christina G., 2013b. An overview of the composition and application of biomass ash.: Part 2. Potential utilisation, technological and ecological advantages and challenges. Fuel 105 (0), 19e39. Verbruggen, N., Hermans, C., Schat, H., 2009. Molecular mechanisms of metal hyperaccumulation in plants. New Phytol. 181 (4), 759e776. Vesterinen, P., 2003. Wood ash Recycling State of the Art in Finland and Sweden. VTT Processes. Research Report Number PRO2/6107/03, PRO2-6107-03, 2003, €skyl€ unpublished results, Jyva a. Ward, C.R., French, D., 2006. Determination of glass content and estimation of glass composition in fly ash using quantitative X-ray diffractometry. Fuel 85 (16), 2268e2277.
Contents lists available at ScienceDirect
Applied Geochemistry journal homepage: www.elsevier.com/locate/apgeochem
Bottom ash of trees from Cameroon as fertilizer €lle Trouve c, Reto Giere d Christoph Maschowski a, *, Marie Claudine Zangna b, Gwenae €t, 79104 Freiburg, Germany Institut für Geo und Umweltnaturwissenschaften, Albert-Ludwigs-Universita ACK e Alumni Club Kamerun, Yaound e, Cameroon c Laboratoire Gestion des Risques et Environnement, Universit e de Haute-Alsace, 68093 Mulhouse Cedex, France d Department of Earth and Environmental Science, University of Pennsylvania, Philadelphia, PA 19104-6316, USA a
b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 27 April 2016 Received in revised form 4 July 2016 Accepted 6 July 2016 Available online 9 July 2016
Utilization of wood bottom ash as fertilizer additive contributes to the return of valuable nutrients to agricultural soils, especially when no artificial mineral fertilizer is being used. In general, wood combustion ash is enriched in calcium and potash, and may also contain elevated amounts of zinc, but the concentrations of these elements depend on tree species, part of the tree, harvest season and local soil type. In this study, bottom ash samples from eight different agricultural wood species from Cameroon, Africa were investigated by using X-ray diffraction and atomic absorption spectroscopy to determine the refractory components and the concentrations of selected heavy metals and arsenic. Results show calcite, potassium salts, periclase and quartz as major components. These phase contents were used to calculate major element concentrations, which were subsequently validated by X-ray-fluorescence analysis. The chemical compositions varied within the range of common compositions of wood ashes. Six of the ashes reached sufficient concentrations of calcium to be defined as a “calcium fertilizer”. Pb contents are most variable, ranging from 0.03 to 21.1 mg/kg. Concentrations of Ni, Cu, Zn, Cd, Pb, and As are all lower than the strictest limit concentrations required for wood ash fertilizers and therefore, the studied wood ashes can be used without environmental concern. © 2016 Elsevier Ltd. All rights reserved.
Keywords: Biomass Wood ash Fertilizer Soil amendment
1. Introduction Solid biomass, e.g. wood, contains various inorganic components in different amounts and speciation, depending on plant type, plant part, soil, and harvest season (Vassilev et al., 2010). Inorganic constituents remain as ash after combustion, as they are non-combustible. Depending on the level of completeness of the combustion, the ash may contain some organic compounds (Vassilev and Vassileva, 1997). The ash is separated into two fractions: one remains on or below the combustion grate as bottom ash; the other is transported along with the flue gas as fly ash, and it comprises the minor part in open fireplaces. Bottom ashes are considered to be unproblematic and sometimes are used as fertil€ ser, izer additives, because they contain valuable nutrients (Ro 2008). Fly ashes, however, may be enriched in (toxic) heavy metals and metalloids and therefore, have to be disposed as problematic waste, after being trapped by a flue gas treatment device. The suitability of bottom ashes for utilization as fertilizer
* Corresponding author. E-mail address: [email protected] (C. Maschowski). http://dx.doi.org/10.1016/j.apgeochem.2016.07.002 0883-2927/© 2016 Elsevier Ltd. All rights reserved.
additive/soil amendment strongly depends on fuel type and soil quality, and has to be assessed for each species and locality (Vassilev et al., 2013a,b). Therefore, the characterization of a wide range of biomass ash types can help in understanding the nutrient cycle and trace element uptake processes, especially when it is crucial to return ashes from biomass combustion to the soil in order to avoid depletion of nutrients. Within the living memory, potash is known as vital nutrient for plant growth, as it consists essentially of water-soluble potassiumbearing salts, which are completely bioavailable (Sharifi et al., 2013). Because the potassium content in wood ash can be as high as 50% by weight (calculated as K2O), this type of ash is being used as a source of potassium for the production of fertilizers, glass (Tite et al., 2006) and soap (Farmer, 2013). However, in modern agriculture, ash from wood combustion is no longer considered a feasible source of potassium, but wood ash might be the only potassium source, when no artificial/mineral fertilizer is used and the soil type is depleted in this element, e.g. when no clay-minerals are present (Barre et al., 2007). Ash from wood combustion is depleted in nitrogen and there€ser, 2008), but can be fore, cannot be used as a general fertilizer (Ro
C. Maschowski et al. / Applied Geochemistry 72 (2016) 88e96
taken as a special amendment to deliver alkaline earth metals, which are able to neutralize acidification of the soil caused by acid rain, i.e. atmospheric pollution (Clarholm, 1994). In general, the calcium concentrations in wood ash are relatively high (Ca mostly present in the form of calcium carbonate and/or calcium oxide), and therefore, wood ash may be used as liming agent to neutralize acid soils and increase the activity of soil bacteria (Bache and Sharp, 1976). In addition, some wood ash types contain enough sulfur and phosphorous that they could be considered as a source of these macro-nutrients. In this study, samples of wood bottom ash from eight different agricultural tree types from Cameroon were investigated by using X-ray diffraction (XRD) with subsequent Rietveld refinement, and by X-ray fluorescence (XRF) and atomic absorption spectroscopy (AAS). The main goals of this study were to identify the crystalline phases and their relative proportions, to evaluate the suitability of combining different analytical techniques, and to compare the concentrations of some troublesome heavy metals and arsenic against strictest limit values for wood ash and other mineral fertilizers (Anon., 2002) (Vesterinen, 2003) in order to find out whether or not the ashes could be used as fertilizers in Cameroon. 2. Samples and methods All trees were harvested in January 2011 in the bush-region of the village Evondo near the international airport ‘Nsimalen’ of , Cameroon (geo-coordinates: 3.743300, 11.524313). To Yaounde represent a real-life scenario, the wood specimens consisted of a mixture of different tree parts (stem wood with bark and twigs). All tree samples were stored for one year prior to combustion for drying reasons. The soil in this region is classified as a “Ferralsol” according to the Harmonized World Soil Database (IIASA, 2012), which stands for iron- and aluminum-rich soil types, often depleted in nutrients. The geological basement underlying the soil consists of metamorphic rocks, represented by embrechite (metamorphic migmatite) and gneiss (Geological map of Cameroon (BMArchives, 2015)). The ash samples studied here were produced by combustion in open fireplaces of eight tree species (Fig. 1), which were burned in , Cameroon. The ashes represent bottom ash the vicinity of Yaounde and were collected directly after combustion, following cooling. All ash samples were ground in an agate mortar and analyzed by XRD with a Cu-Ka X-ray source, using a high-precision setting (2e75 2q; step size of 0.005 ; 5s/step) on a BRUKER D8 Advance diffractometer at the University of Freiburg, Germany. The raw data were processed with pattern matching software (BRUKER DIFFRAC.EVA) in order to identify the phases present (Fig. 2). For each of the identified phases, a structure file was obtained from the American Mineralogist Crystal Structure Database (Downs and Hall-Wallace, 2003). These structure files were then applied to the datasets in the Rietveld refinement program DIFFRACsuite TOPAS by Bruker ltd. to determine the relative proportions of crystalline phases present (Fig. 3). The amorphous content was determined from background signal (Scarlett and Madsen, 2006). Quality of the data retrieved from XRD and Rietveld refinement depends on various aspects: sample preparation (grain-size distribution, surface roughness), degree of crystallinity (amorphous content, e.g. glassy substances and organic matter), and strategy of the refinement procedure (e.g. (Ward and French, 2006)). Grain-size of the sample depends on the preparation method. In our case, we ground it manually with the help of an agate mortar, which is a working practice, but not as accurate as automated grinding. Surface roughness depends on the technique of distributing the sample into the mold of the sample holder, which was done manually as well.
89
Quantification of the amorphous content is usually achieved by spiking the sample with a specific concentration of a known substance, which is 100 percent crystalline. From this information, Rietveld refinement software tools can easily define the degree of crystallinity of the sample. Because of limited sample quantity, we decided to perform XRD measurements without spiking the sample, so that enough uncontaminated material was kept for subsequent XRF and AAS analyses. The calculation of the amorphous content, however, is important; in our case it was done by adding an artificial one-peak phase representing the background signal and thus the amorphous part of the sample. The uncertainty of this method cannot be determined, but the applied method seems to work quite well for the ashes of this study. The refinement procedures were performed without comprehensive corrections like the compensation of preferred orientation, lattice parameters and micro-strain. The bulk chemical composition of the ash samples was analyzed by SGS Lakefield, a Canadian laboratory with ISO17025 standard specializing in geochemical analysis. The specific XRF-analysis method used for this project was the BORATE FUSION/XRF WHOLE ROCK PACKAGE (GO XRF76V). To obtain accurate data for selected heavy metals and arsenic, the samples were also analyzed by AAS at the University of Freiburg, Germany, using an AAS Vario 6 instrument with graphite tube by Analytik Jena. The samples were first digested with a mixture of nitric acid and hydrochloric acid (aqua regia) under microwave treatment. 3. Results The macroscopic colors of the ashes are relatively light for all tree types (see photographs in Fig. 1). The color descriptions given in Fig. 1 conform to the RAL color matching system used in Europe. The light color reflects the presence of only small amounts of organic matter residue, i.e. the combustion process was quite complete. Therefore, the carbon content was not determined and will not be discussed further in this study. 3.1. Mineralogical composition All ash samples contain a considerable amount of calcite (CaCO3), ranging from 25.8 to 70.7 wt% (percent by weight ¼ g $ 100 g1) (Table 1, Fig. 4). They are further characterized by relatively low contents of arcanite (K2SO4), which ranges from 4.17 wt% (mango tree) to 9.26 wt% (cheesewood). Other potassium phases identified include a chloride (sylvite) and a K-bearing carbonate (fairchildite), which is the most variable component, ranging from undetectable amounts (abachi tree) to 38.8 wt% (African drupe tree). The concentration of free lime (CaO) is unexpectedly low (4.42 wt%); in the ashes produced through combustion of avocado tree and African tulip tree the phase was not even detected by XRD (Table 1). Periclase (MgO) is the only magnesium phase present, and its concentration may be as high as 13.1 wt% (African tulip tree). Similarly, halite (NaCl) is the only sodium-bearing phase, but its concentration is very low, except for the ash derived from the abachi tree (12.3 wt%). The content of quartz (SiO2) is highly variable among the different samples (1.67e21.7 wt%), as is the content of amorphous material (7.49e41.4 wt%). 3.2. Bulk chemical composition 3.2.1. Oxide concentrations derived from XRD data In order to assess the utilization potential and quality of the ashes as fertilizer additives, the concentrations of all phases
90
C. Maschowski et al. / Applied Geochemistry 72 (2016) 88e96
Fig. 1. Names and photographs of the source tree species and corresponding bottom ashes analyzed in this study.
containing either Na, or K, or Mg, or Ca, or S have been taken to calculate the corresponding bulk-oxide concentrations. Calculated concentrations of, Na2O, K2O, MgO, CaO, and SO3 were then compared to the minimum concentrations that a potential fertilizer material has to reach. As reference values, we chose as an example the minimum concentrations required for utilization of other r and Haasnoot, ‘inorganic fertilizers’ in the Netherlands (Sarabe 2012), because they also conform to the generally complex regulations for fertilizers in other countries, such as, Germany (Düngegesetz DüG) and Switzerland (Dünger-Verordnung, DüV). Because of the relatively high content of amorphous constituents in some samples and because XRD and Rietveld refinement can only determine crystalline phases, all calculated oxide concentrations
(Table 2) are minimum values. Six of the eight studied samples contained CaO in calculated concentrations that exceed the minimum concentration limit of r and Haasnoot (2012). The two 25.0 wt%, as suggested by Sarabe exceptions, avocado tree and cheesewood ashes, however, contain significantly higher amounts of amorphous components than the other samples (Table 1). This results in underestimation of oxide contents calculated from the XRD data. The calculated contents of the other oxides are below the minimum values suitable for fertilizer additives for all samples. The Na2O concentrations are very low (~1 wt%) or not detectable, except for the Abachi tree ash (6.50 wt%). Although the calculated K2O concentrations are lower than 25 wt%, they are highly variable and range from 2.49 wt% for
C. Maschowski et al. / Applied Geochemistry 72 (2016) 88e96
91
Fig. 2. XRD patterns with phase identification of mango tree ash and umbrella tree ash. Halite and sylvite were not detected in these two ashes and therefore were not labeled.
abachi tree ash to 19.4 wt% for African drupe tree ash. The SO3 content is low, whereas the MgO content is variable and is highest in African tulip tree ash (13.1 wt%). 3.2.2. X-ray fluorescence (XRF) data To evaluate the bulk-oxide concentrations calculated from the XRD and Rietveld refinement data, the ashes were also analyzed by XRF. The values for loss on ignition (LOI) were unexpectedly high, up to 36 wt% (Table 3). Depending on the tree type, the SiO2 content in the ashes is between 7.37 wt% (African tulip tree) and 29.7 wt% (mango tree), which results in the highest variance compared with all other components (Table 3). In contrast, Al2O3 and Fe2O3 contents are low and fairly uniform in the ashes of all tree types, with maximum concentrations observed for cheesewood ash (Al2O3: 4.39 wt%; Fe2O3: 3.06 wt%). MgO concentrations range from 2.94 wt% (mango tree) to 8.42 wt% (African tulip tree). Contents of CaO are high (up to 39.1 wt% for umbrella tree) and, except for the avocado tree and cheesewood ashes, exceed the minimum concentration for utilization as a calcium fertilizer. Na2O content is below 0.2 wt% for all samples. K2O contents, however, are considerably higher and quite variable, ranging from 5.22 wt% (abachi tree) to 15.6 wt% (cheesewood). P2O5 contents are relatively low, ranging from 0.51 wt% for abachi tree to 2.62 wt% for African tulip tree. Comparison of the XRF data (Table 3) with the data derived from XRD and Rietveld refinement (Table 2) reveals that absolute differences in concentration are mostly less than ±5 wt% (Table 4). The
relative differences, however, are high for some of the ash samples, especially in terms of the Na2O content (e.g. avocado, African tulip and abachi trees), because the absolute concentrations of Na2O are so low (Table 3). Because of these relatively low differences, which are due to analytical deviations (semi-quantitative XRD-results are associated with a relative uncertainty of approximately five percent), the probable compositions of the amorphous phases cannot be predicted. Bulk chemical compositions of the eight tree ash samples were compared to those of ashes from the standard DIN þ wood pellet (DINþ is a combination of the German DIN 51731 standard and the € Austrian ONORM M 7135 standard), combusted in a small-scale private wood pellet boiler (50 kW nominal thermal output), and from common wood chips, combusted in a medium-scale biomass power plant (1.7 MW nominal thermal output) located at St. Peter in the Black Forest, Germany. The wood chips were a mixture of softwood species (mainly spruce) from merchantable wood consisting of branches with a diameter >70 mm. Composition of the wood chip ash lies within the compositional range of the eight tree ash samples studied, whereas the pellet ash is depleted in all majorelement oxides, except for SiO2, whose concentration is about three times higher (Fig. 5). The results from XRF analysis provide similar information about the utilization potential of the ashes as fertilizer additives: all ash samples satisfy the requirement for calcium contents, except for the avocado tree ash and cheesewood ash. This result is consistent with that obtained from the XRD and Rietveld analysis.
92
C. Maschowski et al. / Applied Geochemistry 72 (2016) 88e96
Fig. 3. XRD patterns of mango tree ash and umbrella tree ash with Rietveld refinement applied.
Table 1 Semi-quantitative mineralogical composition (in wt%), as determined by XRD and subsequent Rietveld refinement of the eight tree ashes. Mineral species
Mango tree
African drupe tree
Avocado tree
African tulip tree
Dove-wood
Abachi tree
Umbrella tree
Cheese-wood
Lime (CaO) Calcite (CaCO3) Arcanite (K2SO4) Quartz (SiO2) Fairchildite (K2Ca(CO3)2) Periclase (MgO) Sylvite (KCl) Halite (NaCl) Amorphous
4.42 52.3 4.17 12.9 1.43 3.39 n.d. n.d. 21.3
1.67 31.6 7.14 1.67 38.8 5.63 n.d. n.d. 13.4
n.d. 28.0 7.06 21.7 2.85 n.d. n.d. 1.92 38.4
n.d. 47.6 8.63 6.89 2.55 13.1 1.74 1.09 18.5
1.66 56.3 5.87 6.96 1.14 6.64 7.91 n.d. 13.5
2.50 60.6 4.57 5.33 n.d. 7.25 n.d. 12.3 7.49
1.12 70.7 6.25 3.73 1.59 5.20 n.d. n.d. 11.4
0.82 25.8 9.26 13.7 1.21 6.10 1.70 n.d. 41.4
3.3. Heavy metals and arsenic concentrations Because the contents of the heavy metals Ni, Cu, Zn, Cd, and Pb, and of As in the studied ash samples were below the XRF detection limits, these elements were analyzed by AAS. Data (Table 5) reveal considerable ranges in concentration for Cu, Zn, and Pb. Cu content is as low as 64.6 mg/kg in umbrella tree ash and as high as 230 mg/ kg in dovewood ash. Zn ranges from 58.1 mg/kg in African drupe tree ash to 810 mg/kg in avocado tree ash, which nearly suggests this tree species to be a hyper-accumulator for zinc (Rascio and Navari-Izzo, 2011). Pb content exhibits the greatest variability, ranging from 0.03 mg/kg in African drupe tree ash to 21.1 mg/kg in dovewood ash. There is no qualitative correlation between the concentrations of Cu, Zn, and Pb in the different tree ashes, except for dovewood ash and cheesewood ash, which exhibit relatively high abundances of all three heavy metals. Ni exhibits the smallest compositional variation, ranging from 19.3 mg/kg in umbrella tree ash to 34.9 mg/ kg in cheesewood ash. Cd is not detectable by AAS in five of the eight samples and is very low for the others, with the highest
concentration of 0.85 mg/kg in mango tree ash. As is relatively constant too, ranging from 0.79 mg/kg in African drupe tree ash to 2.31 mg/kg in abachi tree ash. As shown in Fig. 6 and Fig. 7, concentrations of the studied heavy metals and arsenic are all below the limit values, which represent the lowest values selected from the national regulations of different countries. Lowest limit values found were those from Sweden (Anon., 2002) and Finland (Vesterinen, 2003). 4. Discussion X-ray diffraction with subsequent Rietveld refinement appears to be a fast and cost-effective method to analyze ashes from biomass combustion because the crystalline refractory materials, the main components of ash, can be determined fairly easily. Furthermore, it allows for the differentiation of phases, which in turn can help in assessing burning temperature and atmosphere €ller et al., 2015). For example, the low concentration of free (Reinmo lime suggests that the combustion temperature was at the very low end of the temperature range for calcite decomposition, which
C. Maschowski et al. / Applied Geochemistry 72 (2016) 88e96
93
Fig. 4. Relative compositions by weight of mineral species in the studied ash samples, as determined by XRD and subsequent Rietveld refinement.
Table 2 Comparison of selected oxide concentrations (in wt%) in the ash samples (calculated from the contents of crystalline phases obtained from XRD and Rietveld refinement) with r and Haasnoot, 2012)); bold values > Cmin fert.; n.d. ¼ not the minimum concentration limits for utilization as ‘other inorganic fertilizers’, labeled as Cmin fert. (data from (Sarabe detected.
Na2O K2 O MgO CaO SO3
Cmin fert.
Mango tree
African drupe tree
Avocado tree
African tulip tree
Dove-wood
Abachi tree
Umbrella tree
Cheese-wood
50 25 15 25 25
n.d. 2.85 3.39 34.1 1.90
n.d. 19.4 5.63 28.5 3.25
1.02 4.99 n.d. 16.4 3.21
0.58 7.95 13.1 27.2 3.93
n.d. 13.7 6.64 33.5 2.67
6.50 2.49 7.25 36.5 2.08
n.d. 4.04 5.20 41.1 2.84
n.d. 7.69 6.10 15.6 4.21
Table 3 Bulk chemical composition (in wt%) of the studied ash samples, as determined by XRF analysis with the minimum concentration limits for utilization as ‘other inorganic r and Haasnoot, 2012); bold values > Cmin fert. fertilizers’, labeled as Cmin fert. (data from Sarabe Cmin fert. SiO2 Al2O3 Fe2O3 MgO CaO Na2O K2 O P2O5 MnO LOI SUM
15 25 50 25
Mango tree
African drupe tree
Avocado tree
African tulip tree
Dove-wood
Abachi tree
Umbrella tree
Cheese-wood
Vari-ance
29.7 2.75 1.46 2.94 31.0 0.10 5.52 0.81 0.44 23.8 98.8
27.5 1.17 0.83 5.04 26.0 0.04 11.0 1.50 0.06 21.2 94.5
23.5 3.65 2.38 5.57 17.5 0.19 13.9 1.68 0.29 29.5 98.4
7.37 1.33 1.28 8.42 30.3 0.07 13.9 2.62 0.03 31.4 97.0
16.6 2.14 2.39 4.81 31.9 0.16 7.43 2.36 0.68 28.6 97.2
14.0 2.20 1.93 5.24 36.7 0.08 5.22 0.51 0.13 29.4 95.6
7.89 1.07 0.85 4.06 39.1 0.06 8.28 0.83 0.18 36.0 98.4
20.4 4.39 3.06 4.05 18.6 0.18 15.6 1.87 0.09 29.8 98.5
70.8 1.45 0.64 2.59 60.7 <0.04 16.5 0.58 0.05 20.5 2.42
Table 4 Comparison of results from XRD and XRF analyses for selected elements as oxides with values showing the differences in wt%; positive values indicate overestimation and negative values indicate underestimation of concentrations from XRD calculations.
Na2O K2 O MgO CaO
Mango tree
African drupe tree
Avocado tree
African tulip tree
Dove-wood
Abachi tree
Umbrella tree
Cheese-wood
0.10 2.67 0.45 3.08
0.04 8.42 0.59 2.47
0.83 8.91 5.57 1.13
0.51 5.95 4.64 3.06
0.16 6.30 1.83 1.57
6.42 2.73 2.01 0.25
0.06 4.24 1.14 2.02
0.18 7.91 2.05 3.03
takes place between 600 and 850 C (Grapes, 2010), (RodriguezNavarro et al., 2009). This conclusion is supported by the relatively high abundance of calcite in all samples (see Table 1). Another contributing factor to the presence of calcite could be the
reaction of CO2 from ambient air with CaO to form secondary calcite. Calcium, however, is also present in another carbonate, fairchildite. These observations suggest that the leaching behavior or liming potential of the studied ashes is similar to any other
94
C. Maschowski et al. / Applied Geochemistry 72 (2016) 88e96
Fig. 5. Comparison of the composition of the ash samples and other standard biomass ash samples.
Table 5 AAS results showing bulk concentrations (in mg/kg) of selected heavy metals and arsenic in the studied ashes; safety values taken from limit values of Ni, Cu, and As in wood ash as fertilizer from Sweden, Zn, Cd, and Pb from Finland.
Ni Cu Zn Cd Pb As
Safety limits
Mango tree
African drupe tree
Avocado tree
African tulip tree
Dove-wood
Abachi tree
Umbrella tree
Cheese-wood
Detection limit
70 400 1500 3 150 30
22.9 85.3 325 0.85 0.11 1.49
23.3 110 58.1 0.27 0.03 0.79
23.9 136 810 0.48 0.13 1.35
24.2 92.6 70.5 <0.02 7.74 1.26
28.9 230 495 <0.02 21.1 1.55
21.5 126 71.81 <0.02 7.86 2.31
19.3 64.6 136 <0.02 4.93 1.60
34.9 159 328 <0.02 11.3 1.25
0.0376 0.0141 0.0080 0.0046 0.0009 0.0012
Fig. 6. Concentrations of selected heavy metals (Ni, Cu, Zn) in the studied ash samples. All concentrations are lower than the lowest limit values which could be found in the literature (limit values of Ni and Cu in wood ash as fertilizer from Sweden, Zn from Finland).
calcium fertilizer. Concentrations of the potassium-bearing phases arcanite and fairchildite, hence the potash content, are relatively low in all studied ashes, except for the ash from the African drupe tree (Table 1). Therefore, the potassium content is of lesser importance in terms of nutrient supply and has a minor, non-deciding influence on the evaluation of the utilization potential of the ashes. In addition to decomposition and dissociation of mineral species during combustion, some elements (e.g. K, S, Na, Cu, Zn, Pb, and As) tend to become volatile as a function of increasing temperature, whereas others (e.g. Mg, P, Mn, Al, Fe, and Si) are more refractory and tend to accumulate in the bottom ash during combustion at temperatures below 1400 C (Misra et al., 1993). Because Zn is an
important macro-nutrient, its content has to be considered. According to Verbruggen et al. (2009), plants with zinc concentrations >10 mg $ g1 (1 wt%, dry basis) qualify as hyper-accumulators. From our bottom ash data, however, we cannot deduce the absolute Zn concentrations in the plants prior to combustion. As the bottom ash from the avocado tree contains such high levels of zinc, we conclude that the avocado tree might be a hyper-accumulator plant. In this study, no correlation of zinc with the concentrations of other elements could be observed. When the aim is that zinc should act as a nutrient, it is important to achieve a pH lower than 7.3 in the soil to avoid the blocking of zinc uptake by plants (Meyer, 1999). The relatively low Na2O contents in all ashes can be attributed to the volatile character of Na. This result is favorable when
C. Maschowski et al. / Applied Geochemistry 72 (2016) 88e96
95
Fig. 7. Concentrations of selected heavy metals (Cd, Pb) and As in the studied ash samples. All concentrations are lower than the lowest limit values found in the literature (limit values of Cd and Pb in wood ash as fertilizer from Finland, As from Sweden). The limit value of Pb is not shown in this figure, because of graphical reasons e it is one order of magnitude higher (150 mg/kg, Finland) than the Pb content of the ash samples.
considering the fertilizer potential of the ashes, because Na is not wanted in soils in significant amounts as it lowers pH, has negative effects on the growth of Na-sensitive plants and can lead to nutrient deficiencies (Davis et al., 2007). The low SO3 content in the ashes can also be attributed to its volatile character at the temperatures of combustion. It is presumed that most of the initial sulfur in the wood was transformed into potassium sulfate (arcanite) during combustion. These potassium sulfate crystals have probably been enriched in the fly ash fraction. Alternatively, they might be present as micro-crystals in the bottom ash, where they cannot be detected by XRD and thus are ‘hidden’ in the amorphous fraction of the ashes. Aluminum is most likely hosted by an amorphous phase, because no crystalline, Al-bearing compounds (e.g. clay minerals) were detected by XRD. The origin of aluminum in wood ash is often attributed to extraneous grains of clay minerals from soil, but the temperature of thermal decomposition of clay minerals (e.g. 800 C for illite (Araújo et al., 2004)) was not reached in our case, as indicated by the abundance of calcite. Because no clay mineral peaks could be found and the decomposition temperature of clay minerals was not reached, we can conclude that no clay minerals were present in any of the ash samples. The amorphous fraction primarily contains glassy substances, including Al- and Si-bearing compounds, but most likely also secondary micro-crystalline phases, including salts (possibly arcanite), as indicated by scanning electron microscope investigations on other wood ash samples (Maschowski, in prep.). The relatively high level of amorphous material in the cheesewood ash may lead to underestimation of the calcium content in calculations from XRD data. Therefore, it is concluded, that a notable amount of calcium may be located in the amorphous part for this sample. High LOI values are most likely due to high water contents, as wood ashes are in general highly hygroscopic (Martin et al., 2013), and not the result of residual carbon, as indicated by the light color of all ash samples (Fig. 1). Furthermore, decomposition of calcite during LOI determination process may contribute to the weight loss.
The mean crystallite sizes of all phases were in an acceptable range (from 100 nm for arcanite to 500 nm for quartz, approximately) and the goodness of fit (GOF) values range from 6.65 for mango tree ash to 8.33 for umbrella tree ash (Fig. 3). These GOFvalues are not far from five, which represents an optimized refinement (Kniess et al., 2012). The concentration of troublesome trace elements, i.e. As, Pb, Cd, Ni, Zn, and Cu might be controlled by the accumulation potential of each particular tree species, as well as by input through regional atmospheric pollution, contamination during harvesting, transportation, and storage. 5. Conclusions In conclusion, we propose a general applicability in terms of sufficient nutrient supply as calcium fertilizer for the six analyzed tree species ashes with minimum of 25 wt% of CaO. The other two ashes, which do not reach this level (avocado tree ash and cheesewood ash), can be used as well, but they may not fully satisfy the needs when calcium is required, for example, to increase pH in acidic soils. Concentrations of the main elements of environmental or health concern (Ni, Cu, Zn, Cd, Pb, and As) are below the lowest limits and they add up to <1 g/kg in all ashes (Fig. 8), and therefore, we propose that the studied wood bottom ashes are suitable as fertilizer and/or fertilizer amendment to be mixed with compost or with one another for use in the local agriculture and forestry of Cameroon. Calcium speciation in the ashes, where the element occurs mainly as calcite, suggests that the ashes can be applied on moderately acid soils. Limit values of troublesome elements, such as toxic heavy metals and arsenic, are for fertilizer substances only, and therefore, the ashes have to be treated with care just like any other waste material. There are no data for properties of the ashes that might be health-relevant. Although the results depend on the tree species, the uptake and ash content of inorganic substances can vary strongly depending on the soil type of the local growing area, atmospheric pollution, contamination during harvesting, transport,
96
C. Maschowski et al. / Applied Geochemistry 72 (2016) 88e96
Fig. 8. Stacked columns diagram of absolute concentrations of selected heavy metals (Ni, Cu, Zn, Cd, Pb) and As of the ash samples.
and storage. In this study, we have shown an applicable and fast-forward procedure of assessing the reuse potential of bottom ashes obtained through uncontrolled combustion of agricultural tree types in an open fireplace. Acknowledgements This work would not have been possible without the financial support by the Europen Union through its Interreg-IV Program (Oberrhein, project C35 BIOCOMBUST; www.biocombust.eu). References Anon., 2002. Recommendations for the Extraction of Forest Fuel and Compensation € nko €ping, p. 20. Fertilising. National Board of Forestry (Skogsstyrelsen), Jo Araújo, J. H. d., Silva, N. F. d., Acchar, W., Gomes, U.U., 2004. Thermal decomposition of illite. Mater. Res. 7 (2), 359e361. IIASA, F. a, 2012. Harmonized World Soil Database (Version 1.2, FAO/IIASA/ISRIC/ ISSCAS/JRC). Rome, Italy and Laxenburg, Austria: FAO and IIASA. Bache, B.W., Sharp, G.S., 1976. Characterization of mobile aluminium in acid soils. Geoderma 15 (2), 91e101. Barre, P., Velde, B., Abbadie, L., 2007. Dynamic role of “illite-like” clay minerals in temperate soils: facts and hypotheses. Biogeochemistry 82 (1), 77e88. BMArchives, 2015. Geological Map of Cameroon [Online] Available at: [Accessed 21 6 2015]. http://www.bmarchives.org/items/show/100202367. Clarholm, M., 1994. Granulated wood ash and a ‘N-free’ fertilizer to a forest soildeffects on P availability. For. Ecol. Manag. 66 (1e3), 127e136. Davis, J.G., Waskom, R.M., Bauder, T.A., Cardon, G.E., 2007. Managing Sodic Soils. s.l.. Colorado State University Cooperative Extension. Downs, R.T., Hall-Wallace, M., 2003. The American Mineralogist crystal structure database. Am. Mineralogist 88 (1), 247e250. Farmer, R.H., 2013. Chemistry in the Utilization of Wood: Pergamon Series of Monographs on Furniture and Timber. s.l.. Elsevier. Grapes, R., 2010. Pyrometamorphism. s.l.. Springer Science & Business Media. Kniess, C.T., Lima, J.C., Prates, P.B., 2012. The Quantification of Crystalline Phases in Materials: Applications of Rietveld Method. s.l.. INTECH Open Access Publisher. Martin, M., Tritscher, T., Jur anyi, Z., Heringa, M.F., Sierau, B., Weingartner, E., vo ^t, A.S.H., Baltensperger, U., Lohmann, U., 2013. HyChirico, R., Gysel, M., Pre groscopic properties of fresh and aged wood burning particles. J. Aerosol Sci. me 56, 15e29.
Meyer, R.D., 1999. Agricultural Use of Wood Ash in California. s.l.. University of California, Agriculture and Natural Resources. Misra, M.K., Ragland, K.W., Baker, A.J., 1993. Wood ash composition as a function of furnace temperature. Biomass Bioenergy 4 (2), 103e116. Rascio, N., Navari-Izzo, F., 2011. Heavy metal hyperaccumulating plants: how and why do they do it? and what makes them so interesting? Plant Sci. 180 (2), 169e181. € ller, M., Klinger, M., Schreiner, M., Gutte, H., 2015. Relationship between ash Reinmo fusion temperatures of ashes from hard coal, brown coal, and biomass and mineral phases under different atmospheres: a combined FactSage™ computational and network theoretical approach. Fuel 151 (0), 118e123. Rodriguez-Navarro, C., Encarnacion, R.A., Ana, L., Rodriguez-Navarro, A.B., OrtegaHuertas, M., 2009. Thermal decomposition of calcite: mechanisms of formation and textural evolution of CaO nanocrystals. Am. Mineral. 94 (4), 578e593. €ser, D., 2008. Sustainable Use of Forest Biomass for Energy: a Synthesis with Ro Focus on the Baltic and Nordic Region. s.l.. Springer Science & Business Media. r, A., Haasnoot, K., 2012. Recycling of biomass ashes in the Netherlands. Sarabe Proceeding of Conference ASH 2012. Scarlett, N.V.Y., Madsen, I.C., 2006. Quantification of phases with partial or no known crystal structures. Powder Diffr. 21 (04), 278e284. Sharifi, M., et al., 2013. Evaluation of liming properties and potassium bioavailability of three Atlantic Canada wood ash sources. Can. J. Plant Sci. 93 (6), 1209e1216. Tite, M.S., Shortland, A., Maniatis, Y., Kavoussanaki, D., Harris, S.A., 2006. The composition of the soda-rich and mixed alkali plant ashes used in the production of glass. J. Archaeol. Sci. 33 (9), 1284e1292. Vassilev, S.V., Vassileva, C.G., 1997. Geochemistry of coals, coal ashes and combustion wastes from coal-fired power stations. Fuel Process. Technol. 51 (1), 19e45. Vassilev, S.V., Baxter, D., Andersen, L.K., Vassileva, C.G., 2010. An overview of the chemical composition of biomass. Fuel 89 (5), 913e933. Vassilev, S.V., Baxter, David, Andersen, Lars K., Vassileva, Christina G., 2013a. An overview of the composition and application of biomass ash. Part 1. Phaseemineral and chemical composition and classification. Fuel 105 (0), 40e76. Vassilev, S.V., Baxter, David, Andersen, Lars K., Vassileva, Christina G., 2013b. An overview of the composition and application of biomass ash.: Part 2. Potential utilisation, technological and ecological advantages and challenges. Fuel 105 (0), 19e39. Verbruggen, N., Hermans, C., Schat, H., 2009. Molecular mechanisms of metal hyperaccumulation in plants. New Phytol. 181 (4), 759e776. Vesterinen, P., 2003. Wood ash Recycling State of the Art in Finland and Sweden. VTT Processes. Research Report Number PRO2/6107/03, PRO2-6107-03, 2003, €skyl€ unpublished results, Jyva a. Ward, C.R., French, D., 2006. Determination of glass content and estimation of glass composition in fly ash using quantitative X-ray diffractometry. Fuel 85 (16), 2268e2277.