Valorisation of agricultural biomass-ash with CO2

This work is part of a study of different types of plant-based biomass to elucidate their capacity for valorisation via a managed carbonation step involving gaseous carbon dioxide (CO2). The perspectives for broader biomass waste valorisation was reviewed, followed by a proposed closed-loop process for the valorisation of wood in earlier works. The present work newly focusses on combining agricultural biomass with mineralised CO2. Here, the reactivity of selected agricultural biomass ashes with CO2 and their ability to be bound by mineralised carbonate in a hardened product is examined. Three categories of agricultural biomass residues, including shell, fibre and soft peel, were incinerated at 900 ± 25 °C. The biomass ashes were moistened (10% w/w) and moulded into cylindrical samples and exposed to 100% CO2 gas at 50% RH for 24 h, during which they cemented into hardened monolithic products. The calcia in ashes formed a negative relationship with ash yield and the microstructure of the carbonate-cementing phase was distinct and related to the particular biomass feedstock. This work shows that in common with woody biomass residues, carbonated agricultural biomass ash-based monoliths have potential as novel low-carbon construction products.

Challenges/issues from an environmental perspective. Biomass and their residues are low-cost voluminous material resources that are generally environmentally benign, often being returned to the soil as an enriching media. Greater than 2 Gt of unused crop residues are dumped in municipal landfills or burned by households in developing countries 32 . These activities contribute to 18% of total global CO 2 emissions [33][34][35][36] .
The use of biomass waste to produce energy and other products is of mounting interest. In this regard, the energy potential of residues, in 2050, has been estimated to be in a range of 15-280 EJ yr −1 globally [37][38][39][40] . However, this estimation of biomass residues generated or their potential for alternative uses is approximate 41 . As such, the global availability of crop residues has been re-examined by Tripathi et al. (2019) 36 .
In most developed and developing countries, the collection, recycling and sustainable disposal of the increasing quantities of biomass and other solid wastes are the major challenges. Their conversion into energy and other products can reduce environmental harms and generate much needed value, particularly in developing countries with large quantities of available biomass.
By far, the greatest use of biomass wastes has been as an energy source by direct combustion of wood and crop residues 42 and in developing countries, biomass is dominantly used as a fuel in open fires for cooking and Table 1. Global production of fruits, vegetables and residues.  44 . Globally, nearly three billion people rely on biomass-based fuels sourced from wood or charcoal for cooking and heating 45 . However, for biomass-based power generation, the EU and USA account for most 46 . Biomass burned in Europe to produce energy is projected to contribute 20% of the European renewable energy target by 2020 47 . As biomass incineration and pyrolysis generates substantial amounts of ash and CO 2 , innovative management strategies that can incorporate both are timely. Ideally, biomass ash should be returned to land as an enrichment, but energy from waste generates large volumes of ash that fall within waste management regulations, necessitating management by for example, landfilling. These landfill 'deposits' are a relatively consistent potential resource for manufactured products. Globally, the demand for 'carbon efficient' management solutions to minimise CO 2 emissions and utilise waste is increasing 48 . The Paris agreement recommendations are for immediate action to keep the global increase in temperatures below 1.5 °C 49 . As carbon capture and storage (CCS) in the geosphere is slow to mature, there is a need to explore carbon capture and utilisation (CCU) for the management of point source CO 2 emissions. Indeed, emerging CCU technologies offer significant opportunities for the management of biomass waste coupled with value-addition and CO 2 emissions reductions (hence the acronym, CCUS).
As a recent example, in November 2019, in Delhi and its adjacent major state, Punjab, suffered record levels of smog and poor air quality. The major contributor (about 50%) was stubble burning by the farmers. In one single day 5,953 fires burned and a monthly total of 31,267 fires was recorded 50 .
As mentioned, the production of cement has a high associated carbon footprint, as calcination generates large amount of CO 2 gas (approximately 650-750 kg CO 2 /t of cement produced). Some 7% of worldwide greenhouse gas emissions are attributed to cement production 51 .
During 2014-2017, the IEA (2019) 30 reported an annual increase of 0.5% in clinker-to-cement ratio, resulting into an increase of 0.3%/y in the direct CO 2 intensity of cement production. This report emphasised the need for an annual decline in emissions of 0.7% by 2030 and deployment of CCUS-based technologies to achieve a sustainable emissions reduction scenario.
An approach that can combine biomass waste management with a reduction of cement production-related emissions is described below. Thus, the present work involves the transformation of CO 2 into mineral carbonates on biomass ash, in a way previously described for other thermal or mineral wastes 52,53 .

Materials and methods
In the present study, biomass residues are categorised as: shell, fiber and soft peel wastes. The biomass residues derived from shell include cobnut, coconut, walnut, almond and peanut; fiber includes jute (hemp), flax, barley straw, hay, and husks from rice and sugarcane; and soft peel includes sweet lime, orange, banana, yam, cassava, potato and pomegranate. The biomass residues described were sourced from India, Africa and the UK.
The residues were ashed in a muffle furnace at 900 ± 25 °C, over 4 h and then examined for (1) selected physical properties (e.g. particle size, bulk density, surface area and ash content) and (2) chemical composition (total carbon, elemental and phase-chemistry).
The particle size distribution of ashes was measured by laser diffraction analysis (Malvern Mastersizer MS2000) and bulk density by loose compaction in cylindrical holders (expressed as kg/m 3 ). The surface area was determined (Micromeritics Gemini V2.00), and total carbon was analysed by CHN analysis (FLASH EA 1112 Series). The bulk elemental composition was determined by X-ray fluorescence spectrometry (Philips LW1400 and XRFWIN software).
The biomass ashes were moistened (20% w/w, total weight) to examine their reactivity to pure CO 2 at a pressure of ~ 2 bar. The ashes were exposed to CO 2 for four-separate cycles in a closed pressurised carbonation chamber, with the first three cycles extending to one hour each, and the fourth cycle being 24 h. The uptake of CO 2 in ashes was determined on weight gain (% w/w, total weight) basis. This approach was taken as the results obtained correlate closely to those experienced during the carbonation of wastes in commercial facilities 36,54 . Product development and characterisation. For the production of monolithic specimens, biomass ashes were moistened (10% w/w, total weight) using a dropper followed by thorough hand-mixing, before being cast as small monolithic cylinders (7 mm × 7 mm-a similar size to manufactured carbonated aggregates). The casting process involved placing the moist ash in the mould followed by hand tamping and the top surface being struck, using a straight-sided spatula. Cylinders in their moulds were placed in a closed curing chamber containing pure CO 2 at 50% RH. After 1 h, samples were de-moulded, and returned to the CO 2 chamber to complete their cycle of 24 h exposure. Non-carbonated samples were treated similarly, but without exposure to CO 2 and were regularly too fragile to demould after curing in air had been completed.
Some of the biomass ashes (including wood biomass ashes) are discussed in Tripathi and Hills et al. 2020 54 . As mentioned earlier, it was necessary to add Portland cement raw biomass waste to provide a reference point, as Portland cement is a commonly used hydraulic cementitious binder. Raw biomass with and without Portland cement was mixed with fine sand (used as an inert mineral filler to change particle size distribution) and then cast into larger monolithic cylinders (3.4 cm × 3.4 cm) ( Table 6). It should be emphasised that the Portland cement was used here for its ability to react with CO 2 gas and produce calcium carbonate rather than its normal use as a hydraulic medium. Cylinders were cured in pure CO 2 for one week.
Assessment of CO 2 uptake and strength in valorised biomass products. The CO 2 uptake by the monoliths was calculated on weight gain (% w/w, total weight) basis and also by CHN analysis. The strength of these monolithic products was evaluated by applying a force until the cylinders failed. The strength was calculated by using the Eq. (1): www.nature.com/scientificreports/ where σ c is the compressive strength in megapascals, F c is the fracture load in kilonewtons, Am is the mean area of the cylinder, and dm is the mean diameter of the cylinder. For each batch of carbonated cylinders, the average strength was calculated from the load recorded at failure, with the three axes of each cylinder being measured using digital callipers (Mecmesin MFG250).
The water resistance of carbonated 'ash only' monoliths was monitored by immersing them in tap water for 30 days to investigate their water sensitivity.
The biomass ashes and resultant carbonate-cemented products were investigated by X-Ray diffractometry and electron microscopy.
The biomass ashes without and with CO 2 exposure were analysed with a Siemens D500 diffractometer, fitted with a Siemens K710 generator using 40 kV voltage and 40 mA current, between 5° and 65° 2θ. The interpretation of diffractograms was aided by DIFFRAC plus EVA software (Bruker AXS) and Rietveld refinement.
The capture of back-scattered electron micrographs augmented by EDAX analysis (JEOL JSM-5310LV, Oxford Instruments Energy Dispersive Spectrometer-EDAX) was performed on polished resin blocks, for both carbonated and reference (no-carbonated) biomass-ash products.

Results and discussion
By combining biomass residues (both raw and ashed) and CO 2 gas into solid monolithic products a potential future 'zero waste' option for these residues is established. Indeed, 4 individual biomass wastes presented in this work have been recently examined by the authors 54 and are used here as reference residues.
The potential of biomass waste with the right chemical and mineralogical composition to react with gaseous CO 2 is harnessed to develop products that are analogous to those made with hydraulic cement. The manufactured products include those hardened by 'ash only' and where ash was used as a partial replacement for cement. Both approaches were used to encapsulate raw biomass in different combinations.
The potential for further innovation including the integration of a direct flue-gas capture and mineralisation step could have significant environmental benefits, not least as this 'circular economic' approach will reduce both the landfilling of biomass waste/residue and gaseous emissions, whilst protecting virgin resources. An offsetting of CO 2 via the replacement of hydraulic cement will be a further added benefit. This 'offset' is of particular importance as we have seen, the cement industry is growing at the rate of about 2.5% pa generating 39.3 ± 2.4 Gt during the period 1928-2016, with 90% of this evolved since 1990 55 . Physical, chemical and mineralogical characteristics of biomass residues. The physical and chemical characteristics of the biomass residues as received, and their ashes are given in Table 2. The surface area of raw fibres was higher compared to shell and soft peel with the later having the lowest values recorded. The ashes from soft peel wastes have a higher surface porosity than most of the ash generated from fibre-waste. The www.nature.com/scientificreports/ particle size distribution of ashes was specific to the individual biomass feedstock ( Table 2). The size range of ash particle size was 3-39 µm for soft peel, 14-52 µm for shell and 1.21-45 µm for fibre-derived ashes.
The total carbon content of ashes varied with some soft peel waste giving a higher yield. A similar trend was observed for shell ash. The carbon content in all the fibre-derived ashes, except rice husk, was generally less variable and within in the same range ( Table 2).
The oxide equivalent composition of ashes is given in Table 3. Walnut shell, jute (hemp) and sweet lime ashes had the highest CaO content in their respective biomass categories. Calcia is the key mineral indicator of the potential CO 2 reactivity of these ashes. It should be noted that coconut shell ash, which had low ash content (0.3% w/w, total weight) was partially characterised as a potential candidate ash, but due to the lack of sample could not be fully evaluated in this study.
Once manufactured, the carbonated 'ash only' monolithic specimens remained intact after immersion in water for 30 days, with no signs of physical degradation, except for the banana peel ash where some swelling and micro-cracking was observed; this was attributed to incomplete carbonation and the presence of CaO, which hydrated forming portlandite (detected by XRD) led to moisture-induced expansion.
Ash content and CO 2 reactivity. The major and minor elements including Al, Ca, Cl, Fe, O, K, Mg, P, Na, S, Mn, Si and Ti in biomass are the major ash forming elements. These elements are generally found in decreasing order of abundance 56 : 56 reported a negative relationship between calcia and ash content in individual species of wood, but not agricultural biomass where a positive relationship was identified. These observations are not universally applicable to woody biomass examined in our as yet unpublished work (where some positive relationships were found), and for the agricultural biomass examined here (where some negative relationships were observed). Some of the soft peel wastes had both high calcia and ash contents (Tables 2 and 3), indicating biomass waste is complex and heterogeneous in nature. By way of example, citrus fruit peels, including orange and sweet lime, have a 'high' calcia content; an observation also reported by Sweitzer (2018) 57 . The ash yield from citrus peel is higher than other 'low' calcia containing fruit peel -an observation also reported by Vassilev et al. (2017) 56 .
The fibres, jute (hemp) and flax, had low ash and high calcia content. This relationship is contrary to Vassilev et al. (2017) 56 observations and may be attributed to these bast fibres having pectin and calcium ions 'gluing' constituent fibres together 58 . On the other hand, rice and sugarcane husk presented high ash and low calcia contents. Vassilev et al. (2017) 56 reported extremely low CaO contents in rice husk, due to high ash yield; data that supports the negative CaO:ash yield-relationship discussed above.
Some shell-derived ashes were calcia rich, up to 52% w/w (total weight), with walnut as an example (Table 3). However, the ash content of shell waste was low, following the negative relationship previously described 56 . That said, the enrichment pattern is opposite to these previously reported findings that reported calcia depleted ash upon burning, which was not observed in our study. Table 3. Mineralogical composition of agricultural biomass ashes (% w/w, total weight). Note: Over 10% is good for reasonable uptake. www.nature.com/scientificreports/ The relationship observed in our work between ash yield and calcia content can be explained by the observation that a high ash results from combustion of biomass particularly enriched in silica/silicates. The ashes abundant in Ca, Cl, Fe, K, Mg, Na, P, S, chloride, sulphate, carbonate and phosphate tend to be of lower yield -also noted by Vassilev et al. (2013Vassilev et al. ( , 2014 59,60 .
The ash arising from herbaceous biomass has been shown to vary with the part of the plant being combusted 61,62 ; with grains having a lower ash content than their straw 63 . Other factors related to the ash content of agricultural biomass include: 1. the loss of nutrients from plants, as a wash-effect by precipitation or delayed harvest 64 , 2. the type of soil used for growing plants 65 , and 3. the season they are grown in, especially for grass species 61 .
As observed in the present study, some fruit peel, shell and fibre-biomass yielded higher calcia containing ashes (see Table 3); the key mineral indicator of CO 2 reactivity. CO 2 uptake in biomass ash and their products Biomass ashes. As mentioned, CO 2 -reactive biomass ashes represent a potential resource for the manufacture of value-added products. However, in an industrialised setting involving, e.g., energy from biomass waste, ashes will be classed as waste and, therefore, to meet 'end of waste' regulations (and to be legally declared as a product), carbonated ashes must: (1) be 'fit for purpose' by conforming to an agreed specification, (2) be risk managed, and (3) meet a market need.
After each successive cycle of exposure to carbon dioxide gas, a gradual increase in the amount of CO 2 uptake was observed (Fig. 1, Table S1).
The potential of ashes to mineralise CO 2 was calculated theoretically, using the Steinour equation 66 . This equation uses the stoichiometry of an ash (taken from its oxide composition) to predict the maximum possible carbon 'uptake' as a % w/w (total weight). However, it should be noted that this equation and modifications thereof are not appropriate to all potentially carbonate-able wastes and the predictions are normally much more than can be achieved under laboratory/real-world conditions. The theoretical CO 2 -uptakes in w/w (total weight) in shell, fibre and soft peel ashes were 26-45%, 4-55% and 28-49%, respectively (Fig. 2, Table S2). However, the experimental values recorded after 24 h of CO 2 exposure were indeed much less, being 9.9-15.6%, 25-29.4% and 5.6-24.17% w/w (total weight) in shell, fibre and soft peel ashes, respectively (Fig. 2, Table S2).
The crystalline phases observed in biomass ash included calcium oxide (CaO) and portlandite (Ca(OH) 2 ). Calcium oxide and portlandite are the major elements responsible for reaction with CO 2 under appropriate hydration condition. Hydration of CaO to portlandite (Ca(OH) 2 ) is an extremely exothermic reaction (− 104 kJ/ mol) 67 . On exposure to CO 2 , portlandite forms calcium carbonate (CaCO 3 ) and this is similarly exothermic (− 32 kJ/mol) 68 , which liberates the formerly bound water 68,69 . Generally, > 10% w/w (total weight) of CaO in a material is associated with a CO 2 uptake that leads to hardening by self-cementation via calcium carbonate formation. Nam et al. (2012) 66 suggest that as CO 2 becomes imbibed, the reaction is predominantly controlled by the phase-boundary, whereas later on, it is controlled by the diffusion of CO 2 through the surface carbonation reaction product. For fresh Portland cement, carbonation is essentially completed in three distinct phases involving 8 steps 66,70 . With respect to biomass ash, no pre-treatment was required, and the carbonation reaction involved: gaseous CO 2 diffusing and dissolving into the film of moisture present on ash particles, followed by ionisation to HCO 3 − . As the pH of the moisture film (which now contains dissolved CaO) falls, as it is neutralised, CaCO 3 is precipitated on the surface of ash particles, in pore space around and within the relict planty structures preserved in the ash. The formation of carbonate causes cementation of adjacent ash particles and the infilling of void space to produce a hardened monolithic product (see Fig. 3).  www.nature.com/scientificreports/ The observed maximum CO 2 uptake was in ashes with the highest calcia content in the following order (low to high): jute (hemp), sweet lime, orange, banana and cassava peel, almond, walnut and cobnut shell, with a range from 10.43 to 29.45% w/w (total weight) ( Fig. 2; Table S2). The calcia in these ashes was between 10.43 and 45.38% w/w (total weight), respectively (Table 3).
Carbon dioxide mineralisation of biomass ash was confirmed by X-Ray diffractometry by the presence of calcite. Rietveld refinement showed the relative occurrence of the major mineral phases in the biomass ashes and in their carbonated counterparts. As might be expected, the mineralogy of the ashes varied between the different biomass feedstock (Table 3). It should be noted that the mineralogy of the ashes is complex, and many amorphous phases are present. As such, the intensity of X-ray reflections is often lower than those obtained for mineral ashes from inorganic feedstock.
When exposed to CO 2, the ashes contained, for example, calcite and monohydrocalcite (observed for lime peel and nutshell). This clearly indicated that CO 2 had been mineralised, and calcium carbonate was formed within the range: 14-67% w/w (total weight). For the sake of brevity, main phases taken from diffractograms of the 'raw' ash and its carbonated counterpart are presented for each category of biomass examined ( Table 4).
The minor presence of portlandite and/or CaO was noted in some of the carbonated ashes indicating that complete carbonation had not been fully achieved, and that further exposure to CO 2 was required. Some of the peel-derived ashes, such as banana and pomegranate were hygroscopic in nature and this could be attributed to the presence of sylvite (KCl); a phase that absorbs moisture from the air 71 . Portlandite development is also responsible for water sensitivity/expansion (and a relative loss of strength in carbonated monolithic specimens) when partially carbonated samples are immersed in water or exposed to the atmosphere 72 .
The literature has much information on the management of biomass ash and its effect on soil properties [73][74][75] , not least as a way of replenishing essential nutrients and for modifying soil structure/microstructure. The very nature of plants to accumulate metals shows there is a key relationship between soil chemistry/mineralogy and the 'needs' of individual plants. Where calcium is concerned, a plant's ability to accumulate this divalent metal can result in ash that readily carbonates on exposure to CO 2 .
As mentioned, the uptake of CO 2 in the biomass ashes was related to presence of CaO arising from the high ashing temperature of 900 ± 25 °C. The mineralisation of CO 2 in the ashes was aided by their finely divided nature/ high surface area, as noted by Castel et al. (2016) 76 78 reported that CO 2 uptaken by, for example, mixed wood ash blended with coal is largely regulated by particle surface area. However, particle size is not always the limiting factor for the CO 2 up taken, as is seen in our study. The findings of Nam et al. (2012) 66 involving municipal solid waste ash showed the amount of CO 2 sequestered increased as particle size decreased. This may well be valid for ashes with a similar chemistry, but where the amount of calcia varies in a feedstock (as seen in the present work) particle size, and in some cases, surface area may be secondary considerations.
Ash only monoliths. The CO 2 uptake in 'ash-only' monoliths (Table 5) showed that walnut shell, jute (hemp) and sweet lime peel reacted with the most CO 2 in their respective categories. However, as noted, a small amount of residual CaO in banana peel ash caused water sensitivity-related micro-cracking as calcia hydrated to portlandite, with a consequent increase in the volume.
Many of the biomass ashes studied have been shown to be very reactive to carbon dioxide gas forming calcium carbonate. The morphology of these carbonate products was examined using polished sections subject to backscattered electron microscopy. The spatial distribution of carbonate seen in the backscattered electronmicrographs suggests the nature of the individual biomasses may have an influence.
Three of the biomass residues, representing one type from each category studied, are given in Fig. 3a-c to illustrate the distinct morphology and carbonate distribution within the cemented biomass-ash monoliths. The observations made are summarised in each respective figure. % w/w, total wt CO 2 emitted during ashing of b iomass (% w/w, total weight) Predicted (Steinour) CO 2 uptake (based on ash composition) (% w/w, total weight) CO 2 in ash (after 24 hrs of CO 2 exposure) (% w/w, total weight) www.nature.com/scientificreports/ The ashing of biomass at 900 ± 25 °C produced primarily CaO as the main calcium-bearing phase. Other minor phases including quartz and feldspar were also noted, which were present in the raw biomass or formed during ashing.

(a)
The microstructure of carbonated sweet lime peel monolith shows sub-angular grains of ashed material that are relatively uniform in nature, enveloped by groundmass that is generally microporous. Larger distinct isolated spherical voids (typically 3 to 10 µm) are common. EDS shows that calcium is uniformly distributed throughout, individual grains and matrix, however the former tend to be Silica rich and potassium poor. Relict structures originating from the peel were occasionally observed.

(b)
Carbonated almond shell ash appeared well cemented, containing little observable porosity. A small number of spherical pores ca. 2% v/v appeared black in colour, and ranged in size from <5 µm to 100 µm. The matrix appeared darker (i.e. had a lower electron density) with individual angular to sub-rounded grains, (<250 µm) comprising ca. 50% v/v; often containing a lighter-coloured silica rich core, also associated with potassia and alumina. The darker outer portion of grains were Ca rich (an elongate grain, located upper centre left of the micrograph, this 'zoning'.

(c)
The microstructure of jute (hemp)-ash monoliths is unlike the others examined. A high proportion of relict-structures arising from the original biomass were observed which were not broken down during ashing. Small isolated lower porosity patches were observed in the matrix. Larger distinct grains were few, although one spherical grain (upper centre of micrograph) can be easily seen. Generally, calcia was found throughout the sample, but in angular ash particles, typically <125 µm, and more easily identified by their element composition (rather than by BSE), calcia and phosphorous occurred together. www.nature.com/scientificreports/ It was also noted that monohydrocalcite was formed in combination with calcite during carbonation. Monohydrocalcite is a metastable phase which can transform to calcite or aragonite under the right environmental conditions. However, small amounts of phosphate can significantly inhibit transformation, and may explain why some of the samples contained both hydroxylapatite and monohydrocalcite 79 . An example diffractogram showing these phases is given in Fig. 4.
Overall, the biomass ashes investigated were CO 2 reactive, and displayed a self-cementing capacity resulting in a hardened mineralised material. Interestingly, the variations in the microstructure of the mineralised products readily confirms carbonate cementation upon reaction with CO 2 gas. Furthermore, the possibility of achieving carbon balance/neutrality in the encapsulated biomass residues with sequestered CO 2 is established.
Raw biomass and ash monoliths. The embodied carbon calculated in monoliths produced from raw biomass combined with CO 2 -reactive biomass ashes was 10-20% w/w (total weight). For cement-bound composites the amount measured was 10% w/w (total weight), clearly indicating (that despite containing the cement), the monoliths were carbon negative ( Table 6). The mechanical properties of the ash + raw biomass-only monoliths without cement were inferior to those containing cement.
When the CO 2 -reactive biomass ashes were used as a partial replacement for cement (see Table 6), it is possible to obtain respectable strengths and a reduced carbon footprint for the products formed. This particular part of our wider study will be published separately.

Strength of carbonated monolith samples
The carbonated 'ash only' monolithic cylinders were examined for their unconfined compressive strength and density ( Table 5). The strength achieved was greatest for almond shell, straw and sweet lime in their respective categories. As compared to uncarbonated monolithic samples, all the carbonated biomass ashes were stronger. The monoliths made from rice and sugarcane husks, and yam peel recorded the lowest strengths, and this Table 4. Example major calcium containing phases in carbonated biomass ashes (%w/w, total weight) as determined by X-ray diffractometry. Data derived by Rietveld refinement; other phases detected included: periclase, quartz and feldspar.  www.nature.com/scientificreports/ appears to be related to raised silica content (15-65% w/w, total weight). Interestingly, coconut husk ash poorly self-cemented (despite having a modest silica content) and this was attributed to the formation of sylvite (KCl) in the hardened product. In most of the cases, the uptake of CO 2 in ashes corresponded to the strength recorded for the carbonate cemented monoliths (i.e. higher CO 2 uptake = higher strength). However, in some cases, e.g., for banana peel a high CO 2 uptake resulted in a low recorded strength; here being attributed to incomplete carbonation leading to mild expansion and microcracking (Fig. 5).
As a means of comparing how the strengths of the small monolithic specimens made from carbonated biomass ash with commercially manufactured carbonated products reference is made to the European standard for lightweight aggregates 80 . It was found that the strength of all the carbonated ash-monoliths examined exceeded the strength criteria given in this standard, being an average of 0.1 MPa. This strength is also that required for 'end of waste' approval for UK commercially available manufactured carbonated aggregate-products, made from CO 2 reactive inorganic wastes 81 .

Implications
Some of the nutshell, fibre and soft peel ashes have displayed potential to combine with ca. 25% of their own weight of CO 2 . However, most of the fibre-ashes captured much less CO 2 . Nevertheless, the amount of captured CO 2 is the direct offset of the CO 2 emitted during ashing. The development of strength by self-cementation by carbonate formation may not necessarily be sufficient for biomass ash raw-biomass composites to be employed as building materials. Under these circumstances a more 'potent' carbonate-able binder may be used, and Portland cement is one such binder.
As shown in this work, when Portland cement is used as a carbonate-able binder, replacement by biomass ash can be used to increase the embodied carbon in the product whilst maintaining desirable strength characteristics. It is worthy of note from our earlier work 54 , Ca-rich biomass ash from woody feedstock could be used to partially replace Portland cement in a hydraulic system without loss of product strength, providing flexibility in approach.   54 .

Strength of valorised raw biomass and ash products (carbonated) (MPa) CO 2 in valorised raw biomass and ash products (1-week carbonation) (%)
Orange peel + cement (20%) + sand (10%) 400 0. www.nature.com/scientificreports/ The strengths observed for biomass ash monoliths and our reference cement-bound samples ranged between 160 and 450 kPa. Considering that the composites were raw biomass at 70% w/w (total weight), these strengths were enough for the samples to be robustly handled. There is also a considerable amount of embodied carbon in these composite products, ranging between 26 and 45% w/w (total weight).
An emission factor for CO 2 emissions from burning of crop residues has been calculated as 1585 g/kg by Akagi et al. (2011) 82 . However, under the laboratory conditions used in this work, we calculated the emissions from burning of crop residues to be 47-66% w/w (total weight) CO 2 , which is lower than that reported by these authors using a CO 2 equivalent calculation.
The direct offset of CO 2 emissions after carbonation of ashes (i.e. what was mineralised/in the raw biomass) could be calculated based on the CO 2 mineralised the products.
The indirect offset, when Portland cement is used (as a carbonate-able binder), can be calculated through the reduction in use of cement by partial replacement by biomass ash. Further potential benefits include release of land space currently used for dumping biomass residues.
A simplified conceptual diagram shown in Fig. 6 delineates the offset options for CO 2 emission by using our low carbon CCUS approach for the valorisation of selected biomass waste. This diagram, however, does not consider the energy involved in burning biomass, as that is the part of a separate study. Nevertheless, direct and indirect offsets calculated for selected biomass residues amount to 134 Mt of CO 2 /year.

Conclusion
The 'proof of concept' established through this study shows that the residual ash from the burning of certain agricultural biomass waste contains enough calcium oxide (× 100 g/kg), to enable carbonate-hardened monolithic products to be manufactured on exposure to CO 2 gas. A negative relationship between calcia content and ash yield was identified. When fully carbonated, these small monolithic products similar in size to manufactured carbonated aggregate are resistant to water and have acceptable strength, as specified in the European standard for light-weight aggregates, BSEN 13055:2016.
These findings suggest an alternative 'low carbon' route for biomass waste utilisation is potentially available, which can sequester significant (× 100 Mt) amounts of CO 2 in products with value.
We conclude from this study that there are a number of significant potential benefits from utilising biomass waste ash that has been mineralised: • significant amounts of CO 2 can be permanently stored, being up to 29.5% w/w, total weight • processing can be carried under ambient conditions opening up the possibility of using point-source emissions at low cost • ashes readily self-cement and the products have MPa strength, and appear environmentally stable • biomass ash wastes normally disposed to landfill under waste management regulations have a route for valorisation via 'end of waste' • the combining of solid and gaseous wastes in products is a circular economic activity with significant potential sustainability gains • The combined direct and indirect CO 2 offset was calculated as 134 Mt/year for the selected biomass residues (incl. crop, fruit and vegetable waste); this amount is equivalent to, e.g., nearly 40% of the UK's GHG emissions predicted for 2019.