Organic coating on biochar explains its nutrient retention and stimulation of soil fertility

Amending soil with biochar (pyrolized biomass) is suggested as a globally applicable approach to address climate change and soil degradation by carbon sequestration, reducing soil-borne greenhouse-gas emissions and increasing soil nutrient retention. Biochar was shown to promote plant growth, especially when combined with nutrient-rich organic matter, e.g., co-composted biochar. Plant growth promotion was explained by slow release of nutrients, although a mechanistic understanding of nutrient storage in biochar is missing. Here we identify a complex, nutrient-rich organic coating on co-composted biochar that covers the outer and inner (pore) surfaces of biochar particles using high-resolution spectro(micro)scopy and mass spectrometry. Fast field cycling nuclear magnetic resonance, electrochemical analysis and gas adsorption demonstrated that this coating adds hydrophilicity, redox-active moieties, and additional mesoporosity, which strengthens biochar-water interactions and thus enhances nutrient retention. This implies that the functioning of biochar in soil is determined by the formation of an organic coating, rather than biochar surface oxidation, as previously suggested.

spectral decomposition of X-ray absorption spectra that were each normalized to a 1 nm layer and cobtained from different ultrathin-sections of biochar particles. "Pristine core" refers to an ultrathin-section cut from the center of a pristine biochar particle after removing the outside by trimming. "Pristine edge" refers to an ultrathin-section cut from a pristine biochar particle as close to the outer surface as possible. "Comp edge" refers to an ultrathin-section obtained from a co-composted biochar particle as close to the outer surface as possible.
(B) Normalized relative contributions of different functional groups to the X-ray absorption across the C1s absorption edge in ROIs defined according to their N/C ratio as described in Figure 4. These semi-quantitative results are based on spectral decomposition of X-ray absorption spectra that were each normalized to a 1 nm layer and obtained from three regions of interest in different spots; spot 74 is shown in Figure 4 of the main manuscript. All high N/C ROIs showed a lower relative abundance of aromatic. No consistent trends could be described for aliphatic and o-alkyl carbon. In high N/C entire area ROIs, relatively more carboxyl C was found as well as at least 4 times higher abundance of carbonyl C and less phenol C than in particle ROIs. Redox-active quinone C hardly showed a consistent trend, but in one region (72), we found a 6 times higher abundance of quinone C in the high N/C particle ROI. In the spectral decomposition, no consistent trend was found that would indicate a significant presence of carbonates, that would result in an absorption peak at 290.3 eV through the 1s  σ* excitation.
(C) RGB-map showing the distribution of biochar, saccharose and protein, calculated from an image stack across the C1s absorption edge through linear spectral decomposition with the spectra of three reference compounds. Saccharose is a contamination as a result of sample preparation. Protein was used as an exemplary N-rich carbonaceous compound. Scale bar represents 500 nm.
high angle annular dark field) micrographs and EELS (electron energy loss spectroscopy) spectra of ultra-thin sections of co-composted biochar.
(A): Thick organic coating on surface that are most likely located inside a biochar pore as indicated by only thin gold coating that is visible as individual bright spots. Rippled appearance of the biochar is an artefact of the ultrasonic knife used in the ultramicrotom. As the biochar is more brittle than the glue and the coating, it is only the biochar to show this artefact.
(B) Close-up of (a) on a thick (100 nm) and intact part of the coating highlighting its porous nature.
(C): Overview on a partly delaminating coating; delamination is an artefact of the mechanical force applied during microtoming. It shows the three-dimensional nature of the porous coating and that delaminating preferentially separates biochar and coating. (a) STEM HAADF micrographs of an ultra-thin section of pristine biochar. Rippled appearance of the biochar (1) is an artefact of the ultrasonic knife used in the ultramictom. As the biochar is more brittle than the epoxy resin (2), it is only the biochar to show this artefact.
High abundance of gold (3) that is not high enough to form a continuous coating indicates a semi-exposed location of this spot in the original biochar particle. No porous coating could be detected. (4) indicates a region of bright spots that show a high content of Ca according to EDX (data not shown).
(b) Close-up on the biochar matrix show the rippled appearance and the ubiquitous presence of bright dots in the nm range. EDS revealed their higher content of inorganic matter compared to the bulk biochar matrix, including Si, Ca, P, As, Cu. As is an omnipresent tracemetal, that can be present in low concentrations in all kind of biomass.
(c) Close-up on a region of biochar (1) with up to 15 nm of continuous gold coating (3) indicating a more exposed location of this region in the original biochar particle ((2) = resin), which contains a region with high abundance of Ca-nano-hotspots as described in (a). No porous coating detectable.    Table 1) that was shown to be concentrated near the biochar particle surface ( Fig 5E). Ca potentially precipitated as the pH dropped when the washing eluate (0.05 M NaOH, pH 10) was replaced with de-ionized water as a fourth and final washing step.  (B) Total pore volume measured at a P/P0 close to 1, total pore volume applying the QSDFT method, volume due to micropores determined with the QSDFT method, volume due to mesopores up to 34 nm with QSDFT methods and volume for pores bigger than 3 nm with BJH method. Volume due to micropores measured with CO2 adsorption and the NLDFT method. (C) Pore size distribution for micropores and mesopores up to 30 nm in pore width according to the DFT method applied to the N2 (grey) and CO2 (blue) adsorption isotherms.
For N2 adsorption the QSDFT method assuming slit/cylindrical pores was employed, while for CO2 adsorption the NLDFT was applied.
high angle annular dark field) micrographs and EELS (electron energy loss spectroscopy) spectra of ultra-thin sections of a co-composted biochar. A "Kon-Tiki" wood biochar 1 was co-composted in parallel to the biochar discussed in the main manuscript.
All preparatory steps (compost feedstock, aeration, picking of the biochar, preparation of ultrathin sections after coating with gold) were conducted as described above. Biochar (1) shows a rippled appearance due to the ultrasonic diamond knife; Gold coating (2)    Fraction of DOC remaining in the column, implying a strong hydrophobic interaction with the column material, comprising longer chain aliphatic and polycyclic aromatic material.

Index
Aromaticity Aromaticity provides an estimation of the degree of aromatic and unsaturated structures of the humic fraction.
Molecular weight A derived value of average molecular mass of the humic fraction.
Inorganic colloids Negatively charged inorganic polyelectrolytes, polyhydroxides and oxyhydroxides of Fe, Al, S or Si, detected by UV light-scattering.

SUVA
An additional parameter derived from the ratio of DOC and spectral absorption coefficient.

Supplementary Discussion
Adsorption and desorption isotherms for N2 and adsorption isotherms for CO2 differ strongly from a quantitative perspective. It is important to highlight that the desorption isotherms in the N2 measurements follow almost a horizontal trajectory, i.e., all N2 isotherms present an open loop hysteresis, which may suggest that equilibrium has not been reached, leading potentially to an underestimation of the specific surface area (SSA) and pore volume (PV). This phenomenon has been previously presented in literature 4 and is mainly caused by diffusion limitations in the pore network, which are significant when the porous structure is a complex network of mostly micropores (<2 nm), with a significant contribution of small (<1 nm) and/or constricted micropores. Such diffusion limitations are mostly observed with N2 adsorption due to the low measurement temperature (77 K). Other phenomena such as "swelling of a nonrigid porous structure or irreversible uptake of molecules in pores (or through pore entrances) of about the same width as that of the adsorbate molecule" 5 could also contribute; these phenomena were also reported by Ravikovitch and colleagues 4 .
For the above-mentioned reasons these measurements were combined with CO2 adsorption, in order to gain more information about the porous structure. Due to the higher temperatures (273 K vs. 77 K of N2 adsorption) and higher absolute pressures of CO2 measurements, the diffusion of CO2 molecules to/into the pores is much faster, even for small micropores, making it possible to characterize a part of the pore network that is potentially not being characterized by N2 adsorption. The disadvantage of using CO2 is that only micropores up to 1.4 nm in pore width can be measured. This is confirmed by the results shown in Figure 8, where the SSA obtained with CO2 adsorption (measuring pores up to 1.4 nm in pore width) is significantly higher for all biochar samples than the SSA obtained with N2 adsorption. The results for N2 adsorption presented in Supplementary Fig. 13B were determined applying the methods previously introduced on adsorption isotherms. Total SSA according to the BET method, total SSA, SSA due to pores <2 nm (~1 nm < pore width < 2 nm) and SSA from pores > 2 nm (2 nm < pore width < ~34 nm), according to the QSDFT method, assuming slit/cylindrical pores and SSA for pores > ~3 nm, according to the BJH method were obtained from N2 adsorption. Total SSA for pores <1.4 nm (~0.4 nm < pore width < ~1.4 nm) according to the NLDFT method was obtained from CO2 adsorption. With respect to the pore volume, analogous explanation is valid.
Comparing these results for all biochar samples it is observed that BCprist and BCprist-washed have both the highest total SSA and the highest pore PV, due mainly to micropores. It is also easy to observe that all the samples follow the same qualitative behavior, meaning that those with the highest total SSA or total PV are also those with the highest SSA or PV due to micropores and mesopores (2 nm < pore size < 50 nm). This shows that all biochar samples have similar porous network from a qualitative perspective, i.e. co-compositing leads to decrease of SSA and PV while washing increases SSA and PV, with no considerable change in the porous system, except the decrease or increase of micropores contribution to total SSA and PV, respectively. An exception to this seems to be BCprist-washed. It would be expected that this biochar presents higher SSA and PV than BCprist due to the washing procedure (e.g. removal of pyrogenic matter that is not part of the actual biochar matrix 6  be much more flexible, allowing CO2 diffusion in the pores. In the present study, the molecules (both organic and inorganic) would not be on the outer layer of the porous structure, but already inside the pores, in their way out of the biochar particle due to the washing procedure, but the impact could be similar. With respect to Supplementary Fig. 13C it is also important to clarify that SSA and PV values from N2 adsorption of pores between 1.25 and 1.45 nm are potentially the sum of SSA and PV of these pores and also pores <1.25 nm, which due to the aforementioned N2 limitations cannot be filled completely filled by N2 molecules but still contribute to some extent. This explains the high values correspondent to this pore width. The QSDFT method applied in the relative pressures of the present study does not show the contribution of pores smaller than 1.25 nm.
Comparing the results further, co-composting of biochar leads to ~18 % reduction in CO2 SSA, while to ~64 % reduction in N2 SSA ( Supplementary Fig. 13C) In the case of CO2, this reduction takes mostly place for pores between 0.65-0.85 nm pore width. For N2, the reduction is for the smallest pore size given by the method (1.25-1.45 nm). The reduction in total SSA is probably a combination of two phenomena: the organic matter of the coating has a lower SSA than the original biochar, hence the reduction in SSA per unit of mass of the BCcomp; and the organic matter may constrict some pores or even block the access, especially for N2 adsorption, explaining the significant differences in reduction of both SSA, similar to the results reported by Pignatello et al. 7 .
Washing of biochar leads to ~10% increase in CO2 SSA for BCprist and ~20% increase in BCcomp. In the first case, this increase was very similar in proportion for pores between 0.45-0.65 nm and pores between 0.65-0.85 nm. For co-composted biochar, the washing increased mostly the pores between 0.65-0.85 nm. This increase could be due to removal of organic matter.
With respect to their capacity of adsorbing NO3 -, the effective radius of this ion is 0.196 nm 8 , but the hydrated ion has a radius of ~0.25 nm 9 , which in principle makes it still possible to be adsorbed even in biochar micropores. Therefore, an increase in SSA, including the smaller micropores (0.65-0.85 nm) could potentially improve NO3adsorption. However, high surface is not enough, since the adsorption mechanism of this ion is more complex than that.
In STEM, the biochar appears as a widely homogenous matrix with a constant contrast in HAADF imaging. Micropores (<2 nm) could not be identified, although they were omnipresent according to gas adsorption measurements.
The water soluble organic component of the co-composted biochar obtained by extraction at 50°C for 24 h has significantly higher contents of both total nitrogen and carbon compared to the eluates of the fresh and soil-aged biochar (Supplementary Table 4).
According The nitrogen content of these large molecular molecules was greater by a factor of 17 (soilaged) and 42 (pristine). The average molecular weight and aromaticity of the eluates from the composted and from the aged sample were similar but higher than those from the fresh sample indicating a difference in the origin of these molecules. Biopolymers, building blocks and LMW neutrals where all much higher for the co-composted sample. There were no detected LMW acids for the aged and the composted sample. The major difference was the very large difference in the negatively charged inorganic polyelectrolytes, polyhydroxides and oxidhydrates of Fe, Al, S or Si, detected by UV light-scattering.