Investigation of Biochars in Terms of Vitamin E Adsorption Capacity

Featured Application

This study provides insights into optimising biochar as a carrier for vitamin E delivery in ruminant nutrition. By demonstrating the correlation between pore size distribution and adsorption capacity, the findings suggest that theoretical models can reduce the number of experimental trials needed to identify effective adsorbents. This approach could improve the efficiency of vitamin E supplementation in animal feed, potentially enhancing nutrient absorption and animal health.

Abstract

Vitamin E is important for ruminants’ health. To increase the rate of vitamin E resorption, the use of a carrier is recommended. One authorised porous feed additive is biochar. Biochar’s adsorption capacity is affected by its pore volume, which is determined, among other factors, by the biomass and the production process applied. For this purpose, the vitamin E adsorption capacity of ten commercial biochars with a varying surface area in the range of 2.6 to 20 nm was investigated. The results of these single-point batch experiments were compared to the theoretical results using a monolayer adsorption model. Our hypothesis was proven, as the theoretical model could predict the experimental adsorption capacity. This generally suggests that the number of trials required to identify optimal adsorbents can be reduced. A high percentage of vitamin E adsorption (>90%) was obtained with a short adsorption time of 10 min using an adsorbent dosage of 15.78 g/L and a vitamin E concentration of 1.70 g/L. The highest correlation of vitamin E adsorption existed for the mesopore class, ranging from 3.22 to 4.03 nm in Barrett–Joyner–Halenda surface area. This indicates the necessity of knowing the size of the adsorptive and the adsorbent in order to optimise sorption kinetics.

1. Introduction

High-yielding milk cows are more prone to a weak immune system, and therefore to infections and illnesses [1], resulting in a higher prevalence of udder infections, such as mastitis [2,3]. As a consequence of udder infections, milk quality and quantity can be affected; among others, the contents of whey, casein, and fat increase, while the contents of lactose and calcium decrease [4,5,6]. Due to these discrepancies, the milk’s production properties are altered [7], resulting in a shorter shelf life and possibly accompanied by negative organoleptic properties [8,9]. To prevent illnesses, it is crucial to keep animals’ health optimal. One pillar of achieving this is the supplementation of an adequate amount of fat-soluble vitamin E [10,11]. The reasons lie in the important effects on metabolism caused by its antioxidative, anti-inflammatory, and immune-modulating effects [12,13]. As a result of rumination, vitamin E is partly metabolised in the rumen, which diminishes its availability and therefore its resorption in the small intestine [14,15]. To transport sufficient vitamin E in the small intestine, the use of a vitamin E carrier is recommended. Adding vitamins or porous substances to feed is common practice to improve animal productivity and health [16].

For instance, zeolite materials have been used as a dry delivery or antioxidant carrier system to transport vitamin E in humans [17]. Available carriers for ruminant feed include mineral silicates, such as zeolites [18], which occur naturally, but can also be produced synthetically for specific purposes [19]. Zeolites belong to the class of crystalline aluminosilicates, due to their rigid anionic framework, which has defined channels intersecting in cavities. Their specific inner surface is 600 to 700 m²/g, with a homogenous distribution of pores with a specific diameter, allowing them to be used as molecular sieves [19]. Due to the synthetization of specific pore diameters, zeolites can be used for specific ion exchange, molecular sieving, or adsorption [18,20]. However, zeolite has been shown to be destroyed in the rumen due to its lack of pH resistance [21]. Bentonite, a clay mineral with a three-layer structure [22], can be applied to dairy cows to adsorb aflatoxin. However, bentonite has also been shown to interfere in the rumen metabolome, indicating possible consequences on cows’ metabolism [23].

The molecular size of the adsorptive defines the minimum pore diameter required for effective adsorption [24]. To effectively adsorb vitamin E, the porous carrier should have a pore size diameter larger than the volume of the vitamin E molecule. Sarker et al. [25] calculated that α-tocopherol, a form of vitamin E, has a size of 1.19 × 0.95 nm, indicating that this two-dimensional shape would fit in micropores with a pore size smaller than 2 nm [19,26]. This assumption is supported by Quinn [27], who determined that the surface area of vitamin E is 0.52 nm²/molecule, and by Maggio et al. [28], who showed that α-tocopherol containing four to nine isopentane units has a surface area of 0.65 nm²/molecule. In their study, Kovalenko and Kuznetsova [29] stated that the vitamin E molecule has a diameter of 26 Å, which corresponds to 2.6 nm.

Besides pore size, the shape of the adsorbent also impacts its transportation and delivery in biological systems [30,31]. An adsorbent with a specific homogenous diameter, which is the case for, e.g., molecular sieves and zeolite [32], can pose a challenge for the adsorption of substances with the possibility of a varying molecular arrangement, such as vitamin E [25,33]. Activated carbons have been described to adsorb bulky organic molecules, such as vitamin E [34]. A comparison of the adsorption capacity of activated carbon and two mesoporous carbon molecular sieves prepared from silica templates indicated that the activated carbon adsorbed less vitamin E than the mesoporous molecular sieves. The reason could be the high quantity of micropores in the activated carbon, which were not filled during the adsorption process [34].

The use of biochar as a carrier for vitamin E is particularly interesting, due to its synergistic effects: biochar can be filled with the vitamin [34], transport it into the animal, and, after desorption, remove unwanted substances to improve animal health [35,36], while the vitamin E can be absorbed. After absorption, vitamin E can contribute to improving udder health through its positive effects on metabolism [11,12], e.g., by reducing clinical mastitis [37,38]. Moreover, biochar can function as an adsorbent of pathogens, improving gut fermentation, which results in an overall improvement in the animals’ immune system [39]. This is why biochar can also function as an alternative to antibiotics [16,40]. A carrier in this field needs to have the following characteristics: (1) it should have a high adsorption capacity for the substance to be adsorbed (vitamin E); (2) the biological activity of the adsorptive (vitamin E) should be kept after adsorption; and (3) the connection between the adsorbent (biochar) and the adsorbate (vitamin E) needs to loosen over time to let the adsorbate desorb [29]. Based on previous studies using carbon-based adsorbents to adsorb vitamin E [25,34], and based on the statement that adsorbed α-tocopheryl acetate retains biological activity [29], we assumed that biochars can adsorb and carry vitamin E, especially α-tocopheryl acetate. As micropores are not accessible during digestion [35], the adsorption of α-tocopheryl acetate in the mesopores is relevant for the use of biochar as a carrier.

Thus, ten biochars, authorised as feed additives [41,42,43], were analysed in terms of their theoretical and experimental capacity to adsorb vitamin E. Using authorised biochars emphasises the application-oriented approach of this study, since the results can directly be transferred to an industrial application. Our results should support the evaluation of a biochar for application as a carrier for α-tocopheryl acetate for transport to the small intestine of dairy cows. The use of biochar as a carrier should prevent the metabolization of vitamin E before reaching the small intestine.

To our knowledge, this is the first study comparing the theoretically necessary pore diameter for the adsorption of α-tocopheryl acetate. Additionally, this is the first time that the theoretical adsorption capacity has been used with the aim to predict a suitable biochar for feeding trials. The hypothesis of this study is that commercially available biochars obtain different adsorption capacities caused by the pyrolysis of varying woody biomasses and production parameters. We hypothesise that biochars with a larger surface area in the range of the molecular size of α-tocopheryl acetate adsorb more. In addition, we hypothesise that the theoretical calculation of the adsorption capacity of α-tocopheryl acetate, using the surface area and theoretical molecular size of α-tocopheryl acetate, is a suitable approach to select a compatible biochar. To prove this, we compare the theoretical with the experimental adsorption capacity.

2. Materials and Methods

2.1. Selection of Biochars

Biochars were selected based on their authorisation to be used as feed for dairy cows in Germany [41,42,43]. To enable direct practical application, e.g., in the feed industry, another aim was to analyse commercially available biochars. Thus, samples of ten biochars authorised for use as feed additives were provided by the respective producers or distributors. The biochars are coded, as the providers do not wish to have a publicly open comparison of the respective adsorption properties. The reason lies in a possible market distortion, as the biochar market in Germany is small. To our knowledge, we received samples of nearly all biochars authorised as a feed additive in 2019. This procedure unavoidably resulted in the disadvantage of not knowing the actual production parameters, such as the pyrolysis temperature, heating rate, or retention time, or the proportions of the biomass. To keep our approach application-oriented, the ten biochars were analysed as received. This means that the particle size distribution and moisture content were neither adjusted nor standardised by sieving or drying. Moreover, biochar should always be administered moist to avoid the formation of dust [39,42]. Ultimately, our decision to proceed with a strongly application-oriented approach by accepting the demands of the providers compromises reproducibility. Based on the selection of biochars authorised as feed additives in Germany, it is clear that the biochars were made of varying woody biomasses and generated under a range of pyrolysis conditions (Table 1).

The analysed biochars were produced with an industrial-scale pyrolysis process and, thus, were (semi-)continuously under controlled conditions. One difference between the patented ‘PYREG’ machine and the kiln is that in the ‘PYREG’, a screw pyrolyzer and a low-oxygen burner are used, and some heat, generated during the burning of pyrolysis gases, is reused to maintain the process. These pyrolysis gases cannot condense on the biochar, which prevents possible negative effects [44]. In general, pyrolysis processes operate at more than 400 °C, whereby the retention time and highest pyrolysis temperature vary among producers [44].

The differentiation between soft- and hardwoods was conducted according to Wiedenhoeft [45]. Hardwoods come from angiosperms, meaning flowering plants, whereas softwoods come from gymnosperms, which are mostly conifers. In the northern hemisphere, most softwoods are needle-leaved and evergreen, such as pine and spruce, while most hardwoods are broad-leaved and deciduous, such as maple, oak, and birch [45].

Table 1. An overview of the pyrolysis processes and biomasses of the analysed biochars. The biomasses were sorted by wood type according to Wiedenhoeft [45].

Biochar	Pyrolysis Process	Biomass	Wood Type
A	Kiln	Mostly beech, some oak and acacia (with very little bark)	Hard
B	PYREG	Sieved fresh wood chips (mainly trunk wood of poplars)	Probably hard
C	Pyrolysis, N/A	Beech	Hard
D	Pyrolysis, N/A	Beech, spruce	Hard, soft (spruce)
E	PYREG	Wood chips	N/A
F	Kiln	Beech, larch, spruce, oak, herbal extracts	Hard, soft (spruce)
G	PYREG	Beech, oak, ash, maple, hornbeam	Hard
H	PYREG	Wood chips, beech wood chips	Hard
I	PYREG	Spruce, pine	Soft
J	PYREG	Mostly beech, some oak	Hard

N/A indicates that no further information was provided.

Table 1. An overview of the pyrolysis processes and biomasses of the analysed biochars. The biomasses were sorted by wood type according to Wiedenhoeft [45].

Biochar	Pyrolysis Process	Biomass	Wood Type
A	Kiln	Mostly beech, some oak and acacia (with very little bark)	Hard
B	PYREG	Sieved fresh wood chips (mainly trunk wood of poplars)	Probably hard
C	Pyrolysis, N/A	Beech	Hard
D	Pyrolysis, N/A	Beech, spruce	Hard, soft (spruce)
E	PYREG	Wood chips	N/A
F	Kiln	Beech, larch, spruce, oak, herbal extracts	Hard, soft (spruce)
G	PYREG	Beech, oak, ash, maple, hornbeam	Hard
H	PYREG	Wood chips, beech wood chips	Hard
I	PYREG	Spruce, pine	Soft
J	PYREG	Mostly beech, some oak	Hard

N/A indicates that no further information was provided.

2.2. Imaging

Biochars were photographed individually with a D5300 (NIKON Corp., Tokyo, Japan) in a photobox equipped with an LED light (HAVOX^® HPB-40D STUDIO PHOTO, Avolux SAS, Vendôme, France) and a ruler as a scale. About 5 mg of each biochar was spread on white paper to illustrate the particle size distribution. The distance between the biochar and the camera was 30 cm, the camera was set to automatic mode, and the zoom was 5×. During postproduction in PowerPoint (Microsoft 365, Microsoft Corporation, Redmond, WA, USA), the photo was cropped, a scale with 10 mm was added, the brightness was increased by 20%, the contrast was reduced by 40%, and the sharpness was enhanced by 25% to emphasise the particles more clearly.

To obtain an overview of the biochar’s structure, scanning electron microscopy (SEM) was applied. The samples were stabilised with liquid nitrogen at −196 °C to form ice (K 1250, Emitech SAS, Montigny-le-Bretonneux, France). Then, the samples were warmed for 20 min at −9 °C and analysed under vacuum at 3 × 10^-4 bar (JSM 6460 LV, JEOL, Akishima, Japan).

2.3. Analysis of Mesopore Distribution

The analyses of the surface area and the pore size distribution were performed at AdFiS products GmbH (Teterow, Germany), according to the protocols of Müller [46]. To determine the Brunauer–Emmet–Teller (BET) surface area and porosity, a surface area and porosity analyser (Gemini VII 2390p, Micromeritics, Unterschleißheim, Germany) was used, measuring the adsorption isotherm of nitrogen at 77 K. Therefore, 0.2 to 0.5 g was degassed under a nitrogen flow before analysis: firstly, for 10 min at room temperature; secondly, for 1 h at 300 °C; and thirdly, cooled to room temperature under nitrogen after degassing [46]. The surface area was deduced from the multipoint BET calculation method using the BET expression in the linear form [47] in the relative pressure range of p/p₀ = 0.05–0.20. To classify the mesopore sizes, despite the IUPAC classification, where mesopores are >2 nm and <50 nm [19,26], mesopores were redefined from > 2.6 nm to < 20 nm, divided into six classes: 2.58–3.22, 3.22–4.03, 4.04–5.16, 5.16–6.92, 6.92–10.26, and 10.26–20.01 nm [46]. The volume in each class was determined by the adsorption of nitrogen and calculated according to the Barrett–Joyner–Halenda (BJH) method [48]. The analysis was carried out in duplicate (n = 2).

2.4. Theoretical Calculation of α-Tocopheryl Acetate’s Volume

Based on the short adsorption time of 10 min (s. 2.6.2), and on previous studies analysing the adsorption of vitamin E by adsorbents [17,25,29,34,49,50], we assumed that the adsorption would take place as linear monolayer adsorption [51]. We searched the literature for the two-dimensional size of vitamin E; however, no information on the size of α-tocopheryl acetate was found. We decided to use the surface area reported by Sarker et al. [25], who also analysed the adsorption of vitamin E. Sarker et al. [25] used the Gaussian 09 suite programme to calculate the two-dimensional shape of α-tocopherol, which was optimised by density functional theory and resulted in a surface area of 1.1305 nm².

2.5. Theoretical Calculation of Adsorption Capacity

The theoretical calculation of the adsorption capacity of α-tocopheryl acetate was conducted using the Langmuir equation, with the following assumptions: (1) monolayer adsorption, meaning that the surface has limited space; (2) the molecule does not alter the occupied surface area; and (3) there is no interaction between the adsorbate and the adsorption of further molecules [51].

Since vitamin E adsorption occurs in the mesopores [34,49], instead of using the BET surface ($S_{BET}$) to determine the theoretical monomolecular and maximal adsorption capacity of α-tocopheryl acetate ($q_{theo,mono,max}$), the results of the BJH surface area ($S_{BJH}$) for each mesopore class, derived from the BJH method, were utilised (Equation (1)). The BET surface area ($S_{BET}$) is not a good indicator for the adsorption of the vitamin E, unlike the BJH surface area ($S_{BJH}$) [25,34]. Depending on the pore surface area ($S_{BJH}$), the theoretical adsorption capacity will vary while other parameters are kept constant. The calculation of the theoretical adsorption capacity was conducted in duplicate, as it was based on each BJH result (n = 2).

$$q_{theo,mono,max}={\displaystyle \frac{{M\times S}_{BJH}\times 1000}{A_M\times N_A}}$$

(1)

-

$q_{theo,mono,max}$: adsorption capacity of α-tocopheryl acetate (mg/g);

-

$M$: molar mass (=472.754 g/mol);

-

$S_{BJH}$: BJH surface area (m²/g);

-

1000: conversion factor;

-

$A_M$: molecular surface area of α-tocopherol (=1.1305 × 10⁻¹⁸ m²);

-

$N_A$: Avogadro number (=6.022 × 10²³ mol⁻¹).

2.6. Adsorption Capacity of Biochars for α-Tocopheryl Acetate

When referring to vitamin E, the most active form, α-tocopherol, is typically implied. Due to the poor storage stability and low availability of α-tocopherol, the most widely used source of vitamin E in animal nutrition is the DL-stereoisomer of α-tocopherol, the synthetic acetate ester DL-α-tocopheryl acetate [52]. α-tocopheryl acetate protects the phenolic group from oxidation caused by light, heat, or oxygen [53]. In contrast to natural α-tocopherol, α-tocopheryl acetate only acts within the animal body, and not in the feed fat, as the antioxidant effect is only released after enzymatic removal of the acetate residue [54]. As a vitamin E derivative, DL-α-tocopheryl acetate (C₃₁H₅₂O₃) (CAS: 7695-91-2, BASF, Ludwigshafen, Germany), with a molar mass of 472.8 g/mol and content of at least 94%, was used.

To transport the fat-soluble vitamin E to the pores for subsequent adsorption, a solvent is necessary, which also affects the adsorption capacity of the adsorbent [29,34,49]. Studies have shown that non-polar n-heptane, polar n-butanol [29,34], or polar ethanol can be used [25,49]. Chandrasekar et al. [49] reported that the vitamin E uptake from n-heptane is higher than that from n-butanol or ethanol, since the latter is hydrophilic. This results in a lower adsorption of the solvent to the adsorbent, which is due to the fact that the competition with the adsorptive vitamin E is lower. On the contrary, polar solvents form hydrogen bonds between the active hydroxyl ether group of vitamin E and the hydroxy group of the solvent. Hartmann et al. [34] declared that even though the vitamin E uptake from n-heptane is significantly higher, and generally, non-polar solvents are required to adsorb large amounts of vitamin E, an ethanol solution is more feasible in medical applications [49]. Thus, ethanol (CAS: 64-17-5, absolute, Merck KGaA, Darmstadt, Germany) was used to transport α-tocopheryl acetate into the pores of the biochar.

As vitamin E is a photosensitive vitamin, during the procedure, the beakers were wrapped in aluminium foil to prevent photodegradation due to light exposure. In addition, the beakers were closed with screw caps. The analysis was carried out in triplicate.

2.6.1. Preparation of Stock Solution

A stock solution of α-tocopheryl acetate with 10.0520 g α-tocopheryl acetate and 9.9641 g ethanol (c = 50.22%) was mixed for 1 h on a laboratory shaker (GFL-3016, Lauda DR. R. WOBSER GMBH & CO KG, Lauda-Königshofen, Germany). From this stock solution, a diluted stock solution (c = 0.22%) was prepared. The diluted stock solution was prepared using 0.87 g, equivalent to 1 mL, of the stock solution, and 199.13 g of ethanol. For homogeneous mixing, the diluted stock solution was placed on a laboratory shaker for 10 min. After each use, the stock solution was stored in a freezer at −18 °C.

2.6.2. Preparation of Biochars with α-Tocopheryl Acetate

The biochars to be analysed were used in their as-supplied condition. Thus, the moisture and particle size distribution differed (see Section 2.1).

To analyse the adsorption capacity of different biochars of α-tocopheryl acetate, 4 g biochar ($m_{biochar}$) was weighed into a beaker, and 10 g diluted stock solution (with a volume of reaction $V_R$ of 0.013 l) was added. The sealed beaker was placed on a laboratory shaker for 10 min to ensure better adsorption of the α-tocopheryl acetate into the biochar pores. The procedure was repeated three times. Until analysis, the samples were sealed, protected from light, and stored in a freezer at −18 °C.

Prior to analysis, testing was performed to determine whether an adsorption time of 10 min was sufficient. Among others, two activated charcoals were used to determine the adsorption capacity of α-tocopheryl acetate using the described procedure. Both activated charcoals had an adsorption capacity of 98.82%.

The theoretical adsorbent dosage was calculated based on the mass of the adsorbent biochar (4 g) and the volume of the diluted stock solution. The volume of the diluted stock solution was calculated based on the mass and density of α-tocopheryl acetate and ethanol in the stock solution and in the diluted stock solution, resulting in 0.25334 L. Thus, the adsorbent dosage was 15.78 g/L, whereby the dry matter of the adsorbent was neglected, but considered in the calculation of the experimental adsorption capacity. The concentration of α-tocopheryl acetate in the diluted stock solution was 1.70 g/L.

2.6.3. Analysis of Biochars’ α-Tocopheryl Acetate Adsorption

To create a calibration line, a stock solution (21.1 mg α-tocopheryl acetate in 25 mL ethanol) with 0.844 mg_{α-tocopheryl acetate}/ml_ethanol, as well as an intermediate dilution (0.75 mL stock solution was diluted in 25 mL ethanol) with 0.025 mg_{α-tocopheryl acetate}/ml_ethanol, were prepared. From the intermediate dilution, as well as the stock solution, the following standards were used: 1.266, 2.532, 5.064, 10.128, 20.256, 25.31, 42.2, 84.4, and 168.8 µg_{α-tocopheryl acetate}/ml_ethanol.

To prepare the sample, 1 g was transferred to a cup and centrifuged at 14,000 rpm and 20 °C for 20 min. If necessary, the supernatant was membrane-filtered. Then, 50 μL was pipetted into an amber glass vial, 950 μL of ethanol was added, and the mixture was thoroughly vortexed.

To analyse the α-tocopheryl acetate content in the supernatant, 200 mg of the homogenised supernatant was mixed with 10 mL of ethanol. Then, the samples were treated in an ultrasonic bath (RK510H, Bandelin Sonorex, Berlin, Germany) for 5 min, and were shaken overnight (150/min; SM30, Edmund Bühler GmbH, Bodelshausen, Germany). After centrifugation (3000× g, 5 min; Heraeus Megafuge 16, Thermo Fisher Scientific, Waltham, MA, USA), aliquots of the supernatants were diluted with ethanol to achieve a suitable concentration. After membrane filtration (0.45 μm; Chromafil PET-45/15 MS, Macherey-Nagel, Düren, Germany), samples were analysed by high-performance liquid chromatography (HPLC System Alliance Separations Modul 2695 with a column oven, and Photodiode array detector 2996, both Waters, Milford, CT, USA), equipped with a LiChrospher 100 RP-18 (5 μm, 250 × 4 mm, Merck, Germany). The eluent was 60% methanol and 40% acetonitrile (isocratic), the flow rate was 1 mL/min, the injection volume was 10 μL, the running time was 15 min, the temperature of the column oven was 40 °C, and the photodiode array detector was at 210 nm. The analysis was performed using a chromatography data system with standard calibration at 210 nm (5 points, linear, correlation > 0.998, reference tocopheryl acetate). The analysis was repeated twice for each sample.

The adsorption capacity $q_{abs.}$ of the biochars of α-tocopheryl acetate was determined according to Equation (2) for each processing (n = 3). The concentration of α-tocopheryl acetate in the stock solution was 1.69 ± 0.08 g/L.

$$q_{abs.}={\displaystyle \frac{(C_0-C_t)\times V_R}{m_{biochar}}}$$

(2)

-

$q_{abs.}$: absolute adsorption capacity of α-tocopheryl acetate (mg_{α-tocopheryl acetate}/g_biochar);

-

$C_0$: initial concentration of α-tocopheryl acetate in the stock solution (g/L);

-

$C_t$: concentration of α-tocopheryl acetate in the stock solution (g/L);

-

$V_R$: volume of reaction (L), which is ${\displaystyle \frac{m_{ethanol}}{{\rho }_{ethanol}}}\pm {\displaystyle \frac{m_{\mathsf{\alpha }-\mathrm{t}\mathrm{o}\mathrm{c}\mathrm{o}\mathrm{p}\mathrm{h}\mathrm{e}\mathrm{r}\mathrm{y}\mathrm{l} \mathrm{a}\mathrm{c}\mathrm{e}\mathrm{t}\mathrm{a}\mathrm{t}\mathrm{e}}}{{\rho }_{\mathsf{\alpha }-\mathrm{t}\mathrm{o}\mathrm{c}\mathrm{o}\mathrm{p}\mathrm{h}\mathrm{e}\mathrm{r}\mathrm{y}\mathrm{l} \mathrm{a}\mathrm{c}\mathrm{e}\mathrm{t}\mathrm{a}\mathrm{t}\mathrm{e}}}}$

-

$m_{biochar}$: weight of biochar (g).

As the moisture content varied between the biochars, the adsorption capacity $q_{abs.}$ was also calculated based on the dry matter of the biochar, according to Equation (3), which was adapted from Yang et al. [55]. The dry matter $m_{drymatter}$ was calculated based on the moisture content of the biochar, which was determined by drying according to §64 LFGB L 06.00-3 2004-07, as reported in Witte et al. [56]. The adsorption capacity $q_{abs.}$ was determined in triplicate (n = 3).

$$q_{abs.}={\displaystyle \frac{(C_0-C_t)\times V_R}{m_{biocharDM}}}$$

(3)

-

$q_{abs.}$: the absolute adsorption capacity of α-tocopheryl acetate (mg_{α-tocopheryl acetate}/g_biochar DM);

-

$C_0$: the initial concentration of α-tocopheryl acetate in the stock solution (g/L);

-

$C_t$: the concentration of α-tocopheryl acetate in the stock solution (g/L);

-

$V_R$: the volume of the reaction (L), calculated as ${\displaystyle \frac{m_{ethanol}}{{\rho }_{ethanol}}}\pm {\displaystyle \frac{m_{\mathsf{\alpha }-\mathrm{t}\mathrm{o}\mathrm{c}\mathrm{o}\mathrm{p}\mathrm{h}\mathrm{e}\mathrm{r}\mathrm{y}\mathrm{l} \mathrm{a}\mathrm{c}\mathrm{e}\mathrm{t}\mathrm{a}\mathrm{t}\mathrm{e}}}{{\rho }_{\mathsf{\alpha }-\mathrm{t}\mathrm{o}\mathrm{c}\mathrm{o}\mathrm{p}\mathrm{h}\mathrm{e}\mathrm{r}\mathrm{y}\mathrm{l} \mathrm{a}\mathrm{c}\mathrm{e}\mathrm{t}\mathrm{a}\mathrm{t}\mathrm{e}}}}$;

-

$m_{biocharDM}$: the weight of biochar (g) on a dry matter basis, which was multiplied by $m_{drymatter}$, which is $100-m_{moisture}$.

In addition, the relative adsorption capacity $q_{rel.}$ was calculated according to Müller [46], using Equation (4). The adsorption capacity $q_{rel.}$ was determined in triplicate (n = 3).

$$q_{rel.}={\displaystyle \frac{{(C}_0-C_t)}{C_0}}\times 100$$

(4)

-

$q_{rel.}$: the relative adsorption capacity of α-tocopheryl acetate (%);

-

$100$: factor;

-

$C_0$: the initial concentration of α-tocopheryl acetate in the stock solution (g/L);

-

$C_t$: the concentration of α-tocopheryl acetate in the stock solution (g/L).

2.7. Statistical Analysis

Statistical analyses were conducted using SigmaPlot 15.0. (Systat Software Inc., San Jose, CA, USA). Data for each mesopore class, as well as the theoretical and experimental adsorption capacity of α-tocopheryl acetate, were tested for normal distribution using the Shapiro–Wilk test and the Brown–Forsythe test for equal variance, with α = 0.05. For the experimental adsorption of α-tocopheryl acetate, the tests performed were passed, and the data were consequently analysed for significant differences using One-Way Analysis of Variance (ANOVA) and Tukey’s multiple comparisons test, with α = 0.05. For comparisons of mesopore class and theoretical adsorption capacity between biochars, neither normality testing nor equal variance testing was passed, which is why the data were transformed using logarithms, roots, reciprocals, or squaring. As the data were still not normally distributed (p ≥ 0.05), a Kruskal–Wallis One-Way ANOVA on ranks, with Tukey’s multiple comparison test, was conducted, with α = 0.05. However, the ANOVA on ranks showed no significant differences (p ≥ 0.05), even if large differences existed between the respective results. This might be attributed to n = 2, as the porosity analysis was only conducted in duplicate. Thus, despite the failed normality (p ≥ 0.05) and/or the failed equal variance tests (p ≥ 0.05), a One-Way ANOVA was conducted, showing significant differences between the biochars (p < 0.05). The same procedure was conducted for the comparison of the mesopore classes for each biochar. Significant differences (p < 0.05) between biochars are indicated with small letters (abcdef), whereas significant differences (p < 0.05) between mesopore classes for each biochar are indicated with capital letters (ABCDEF).

Regression was analysed between the BJH surface area and the relative adsorption capacity of α-tocopheryl acetate using the polynomial linear equation f(x) = y₀ + ax. The decision for this regression analysis was based on the assumption of monomolecular linear adsorption. As results, the strength and direction of the linear relationship, the correlation coefficient r, and the coefficient of determination r², as well as the adjusted coefficient of determination r²_adjusted, are shown.

3. Results and Discussions

3.1. Characteristics of Biochars

The biochars exhibited a widely varying particle size distribution (Figure 1). Since some biochars contained wood fragments with a particle size larger than 2 mm, an analysis of the particle size distribution using laser diffraction spectroscopy was not possible. A sieve analysis was not feasible, as some biochars appeared to be moist. A drying step prior to analysis would possibly have reduced the particle size.

Particles varied from powdery to visibly larger elongated pieces (>10 mm); the latter could be seen for biochars A, D, E, G, H, I, and J, whereas biochars B, C, and F were powdery, with particles smaller than 5 mm (see photographic images). The SEM images reveal that these large particles were wood fragments. The SEM images clearly indicate the variation in particle size distribution within each biochar and among biochars. In particular, biochars B, C, D, E, G, H, and I have a heterogeneous particle size distribution. Although the SEM images give limited quantitative information about each sample, especially regarding the vertical dimensions [57], the irregular shape and size of the wood fragments are clearly visible. These fragments could have been a result of the pre-pyrolysis particle size and/or varying destruction during the production process [58]. In contrast, powdery biochars, such as biochar F, are the result of grinding post production [58]. During the thermal conversion from biomass to biochar, the mineral and carbon skeleton remains, preserving the porosity and, thus, the structure of the original biomass. These residual cellular structures contribute to biochar’s microporosity, and can allow the identification of, e.g., woody biomasses [59]. In woody biochar, these macro-cracks generally appear due to the shrinkage stress as the surface decomposes faster than the inner part of the wood [60]. Thus, these macro-cracks are not only related to the biomass, but also to the rate of pyrolysis [61]. The SEM images reveal that the woody biomasses are more or less still intact. For instance, in the SEM images of biochars C, G, I, and J, the channels of the wood particles are visible. The channels of biochar G are app. 50 µm wide, whereas the channels of biochar C, I, and J are app. 10 µm wide. According to the IUPAC classification system, macropores are larger than 50 nm [19,26], which clearly indicates that these channels are former wood cells.

3.1.1. Influence of Processing

Even if the pyrolysis process is similar for different biochars (Table 1), the pyrolysis retention time, the pyrolysis temperature, and the post-treatment (data not available, see Section 2.1) will differ, resulting in different particle sizes. This assumption is based on the statement that biochar’s particle size distribution is influenced by the biomass and its particle size distribution, and the production processes, including the pyrolysis temperature and retention time, as well as the post-processing, such as grinding [58]. In general, the particle size is reduced at a higher pyrolysis temperature [62], and a higher pyrolysis temperature increases the elemental nutrient concentrations and decreases acidic surface groups, which increases the pH [63]. In general, the surface groups and their kinetics also determine the adsorption capacity [46]. However, as the particle size distribution prior to pyrolysis is unclear, the effect of the pyrolysis temperature on the particle size could not be investigated. Panwar et al. [64] separated the pyrolysis process into fast, intermediate, and slow, which were further separated into batch or continuous processes. Batch processes are conducted with a kiln or retorts, whereas continuous processes are conducted in a drum or screw pyrolyzer or a rotary kiln. For the biochars analysed, the retention time of the pyrolysis process is unknown. However, using a screw pyrolyzer, such as the ‘PYREG’, the retention time can be adjusted [44]. When using a kiln, the retention time depends on the dimensions of the kiln, the pyrolysis temperature, and the biomass used. Another difference is the setting of the pyrolysis temperature: in a kiln, temperatures between 400 and 600 °C are used, whereas the incline is 50 °C [44,65]. A screw-type pyrolyzer can operate up to 900 °C [66], and the lowest temperature should be 350 to 400 °C, with a heating rate of 20 to 40 °C/min [67]. Schimmelpfennig and Glaser [44] classified processes based on analysis of 66 procured biochars. They showed that even if the biochars were produced with a similar process, their properties varied. This was also the case for biochars made with the ‘PYREG’. Still, based on the properties for the use of biochar as a soil amendment, they preferred ‘PYREG’ over a kiln [44].

3.1.2. Influence of Biomass

In addition, the biomass composition did vary between all the tested biochars, and can only be compared based on limited information (Table 1): Biochar B, E, and H consist of wood chips. Biochar B and E appear visually similar due to a similar particle size and varying particle shapes. However, biochar H shows a very heterogeneous particle size distribution, and the particles have a stick-like shape. This indicates that the rough information about the initial particle size of the biomass (‘wood chips’) does not lead to any conclusions about the particle shape and size. Neither the wood type nor the woody biomass significantly affects the visual particle size distribution of biochars, as no clear similarities are observed between biochars made from the same biomasses, with the wood cells (channels) of hard- and softwood biochars appearing similar based on SEM.

Due to the varying particle size distribution, we assume that a decreased particle size results in a reduced diffusion path length for the adsorptive, potentially enhancing the efficiency of the adsorption process. Despite the varying particle size distribution, the intention of this study was to analyse the biochars’ adsorption capacity based on their initial properties. Besides variation in particle size distribution, the moisture content, the pH value, the density, and the composition, among others, will differ as the production processes, and especially the biomasses used, vary. These parameters were excluded, as the purpose of this study was to identify a commercially available biochar with a high adsorption capacity of α-tocopheryl acetate without any pre-treatment, in order to optimally enable easy handling using biochar to transport α-tocopheryl acetate into the intestine.

3.2. Mesopore Analysis

When comparing the BJH surface area between the biochars, the mesopore classes within a single biochar are only homogenously distributed for biochars D and E (Figure 2A). On the contrary, when comparing the BJH pore volume of each mesopore class for each biochar, it is striking that the proportions within the respective biochars are somewhat similar (Figure 2B). This indicates that the proportion of the pore volume to the mesopore classes is homogeneous. Nevertheless, some significant (p < 0.05) differences exist between the pore volumes of the mesopore classes. For instance, biochars D, E, and F have a higher pore volume in the higher mesopore classes (10.26 to 20.01 nm), whereas biochars A and G have a higher pore volume in the smaller mesopore classes (2.68 to 3.22 nm). In all classes, biochar H has the lowest volume, and biochar E has the highest volume.

When comparing the total surface area and total pore volume, except for biochars F and A, the other biochars do not obtain one-third of the surface area or the volume of biochars D and E. This underlines the high porosity of biochars D and E, and their possible high adsorption capacity.

3.2.1. Influence of Biomass

In view of the high surface area and mesopore volume of biochars D and E, and in view of the fact that the pore volume increases with increasing mesopore class, the assumption can be made that the biomass prior to pyrolysis might have had a high initial porosity [58]. Moreover, this leads to the conundrum that, on the one hand, the pore volume is higher for higher mesopore classes, which could have been caused by a high initial porosity of the biomass [58] and low pyrolysis temperatures, but, on the other hand, the pyrolysis temperature could have been responsible for a high pore volume [68,69]. Considerably, the range of the mesopore classes varies from 0.64 to 9.75 nm, which consequently results in a higher potential surface area and mesopore volume in, e.g., the mesopore class ranging from 10.26 to 20.01 nm. However, smaller pore classes need to be differentiated more precisely, as these occur due to increased pyrolysis [68,69] and are mandatory for fixation of the adsorptive [32]. This is underlined by the high surface area in the small mesopore classes.

Considering these results with respect to the biomasses (Table 1), no difference between the high surface area or pore volume of biochars D and E and the low pore volume of biochar H is obvious, as these biochars consist of wood chips and/or beech and spruce. Thus, the statement that the pore volume depends on biomass [70,71] does not necessarily apply in this study. These findings underline that the use of softwood does not necessarily result in a larger surface area than the use of hardwood. Biochar I, the only biochar made of softwood, has a rather small pore volume in comparison to the other biochars. This is contrary to Mukome et al.’s [72] statement that the less dense composition of softwood results in more pores and vesicles. Studies have also shown that pyrolyzed hardwood does not necessarily result in a similar pore volume: biochar made from beech had a pore volume of 0.0028 cm³/g and a surface area of 70.2 m²/g [73], whereas biochar made from oak had a pore volume of 0.6011 cm³/g and a specific surface area of 945 m²/g [74], although both biochars were pyrolyzed at 800 °C. In biochars A, G, and J, oak was only pyrolyzed together with beech, which resulted in a medium pore volume in comparison to the other analysed biochars.

3.2.2. Influence of Particle Size Distribution

No direct correlation between the visual particle size distribution (Figure 1) and the surface area or the pore volume of different mesopore classes exists. Noteworthily, the microstructure illustrated by the SEM images has a scale of 500 µm, whereas the mesopores range from 0.00258 to 0.02001 µm in diameter. In sum, the combination of pyrolysis process and wood type (Table 1) does not necessarily influence the surface area or pore volume. In combination with pyrolysis temperature and retention time, the pyrolysis process and biomass may show a connection with the surface area and pore volume, and could alter the adsorption capacity. In general, the higher the pore volume, the more substance can be adsorbed [32]. Based on these results, biochar E should have the highest adsorption capacity of α-tocopheryl acetate, followed by biochar D, whereas biochar H should have the lowest adsorption capacity.

3.3. Theoretical Calculation of Adsorption Capacity

Using Equation (1), the theoretical, monomolecular, and maximal adsorption capacity of α-tocopheryl acetate in each mesopore class (Figure 3), as well as in the total mesopore class of the respective biochars, was calculated (Figure 4). In comparison with the analysis of the mesopore classes (Figure 2), biochar E consistently has the largest surface area and pore volume in each mesopore class and the largest theoretical adsorption capacity. In sum, the connection between the surface area and pore volume in each mesopore class and the theoretical adsorption capacity is high, since the theoretical adsorption capacity is based on the surface area, which consequently affects the pore volume.

3.4. Adsorption Capacity of α-Tocopheryl Acetate

To compare the theoretical calculation of pore filling, the biochars were analysed for their adsorption capacity of α-tocopheryl acetate. The theoretical adsorption capacity was higher for all biochars than the experimental adsorption capacity, with the extent varying greatly (Figure 4). The order of highest to fifth highest adsorption capacity remained the same for both the theoretical and experimental adsorption capacities (E > D > F > A > J). Interestingly, biochars D and E had a high moisture content, but adsorbed the highest amount of the fat-soluble α-tocopheryl acetate. This shows that the distribution of the pores and the surface area and pore volume is more important than the initial moisture content of the biochar. The difference between the experimental and theoretical adsorption capacity might have been due to the short adsorption time of 10 min. This is supported by Sarker et al.’s [25] finding that a longer adsorption time increased the adsorption capacity. Sarker et al. [25] found an adsorption capacity of 96.2 mg_{α-tocopherol}/g_{activated carbon} after incubation at 25 °C for 24 h for an activated carbon with a total pore volume of 0.56 cm³/g. Biochars D and E resulted in experimental adsorption capacities of 7.3 and 7.6 mg_{dl-α-tocopheryl acetate}/g_{biochar DM}, with a mesopore volume of 0.08 and 0.09 cm³/g, after incubation for 10 min at 20 °C, respectively (Figure 2B). The difference in the adsorption between biochars D and E and the activated carbon might have been due to the difference in pore volume, as well as the adsorption time. Yaneva et al. [17] found a maximum achieved experimental equilibrium sorption capacity after 48 h incubation at 19 °C of 9.9 mg_{dl-α-tocopheryl acetate}/g_{zeolite DM} for zeolite with a pore volume of 0.11 cm³/g. This indicates that in our study, firstly, the short adsorption time was enough for sufficient adsorption; and secondly, the use of biochar as a carrier for α-tocopheryl acetate might be more feasible than zeolite.

We assume that, despite the differences between the experimental and the theoretical adsorption capacity, the feasibility of this approach is meaningful. These results indicate that using Equation (1) to determine the theoretical, monomolecular, and maximal adsorption capacity of α-tocopheryl acetate, under the experimental conditions applied, was correct. Using an increased adsorption time, monomolecular multilayer adsorption might occur instead, resulting in the necessity to calculate the theoretical adsorption capacity with mathematical models other than Langmuir’s [17]. Irrespective of this, considering the experimental adsorption capacity for the respective biochar will provide the necessary information to determine the necessary load, e.g., when biochar is used as a carrier for vitamin E in dairy cows. In addition, free pores of biochars can be useful for these animals, e.g., for the adsorption of unwanted substances, such as pathogens or mycotoxins [16,39].

3.4.1. Correlation of Adsorption Capacity and Total Mesopore Volume

Based on the assumption of monomolecular adsorption of α-tocopheryl acetate into the pores, a linear regression analysis was carried out. The total BJH mesopore surface area explained the adsorption capacity by 88.8% (r²_adj = 0.888), with r = 0.949, which underlines the strong positive linear connection between these parameters (Figure 5). Consequently, knowledge of the total mesopore area, depending on the size of the adsorptive, is of huge importance to predict the adsorption process.

When considering the standard deviations, the homogeneity to heterogeneity of the biochars’ particle size distribution becomes obvious. The more homogeneous the particle size distribution appears (Figure 1), the lower the standard deviation of the relative adsorption of α-tocopheryl acetate. Biochars D and E, which have the largest total mesopore volume and highest adsorption capacities, show a heterogeneous particle size distribution due to woody fragments. Biochars D and E also show powdery particles; however, biochars B and C contain more powder and smaller particles, but a smaller total mesopore volume and adsorption capacity. The impact of biochars having a larger amount of small particles, with an assumed smaller diffusion path length for the adsorptive, was not confirmed. Generally, these differences in the adsorption capacity of α-tocopheryl acetate into the pores of biochars underline the impact of pore volume, which is affected by sorption kinetics [46]. We conclude that, in this study, the influence of the biomass on the pore volume was rather small, whereas the influence of processing seems to have been large. Thus, to gain a comprehensive understanding of the sorption kinetics, more information about biochar and its processing is mandatory.

3.4.2. Correlation of Adsorption Capacity and Respective Mesopore Class

To determine which mesopore class is best suited for the adsorption of α-tocopheryl acetate, a regression analysis of the relative adsorption capacity of α-tocopheryl acetate to the BJH surface area of the six mesopore classes was conducted (

Table 2. An overview of the correlation coefficients r, coefficients of determination r², and adjusted coefficients of determination r²_adj, as well as the corresponding function for the Barrett–Joyner–Halenda surface area of each mesopore class, and their dependence on the relative adsorption capacity of α-tocopheryl acetate.

Mesopore Class (nm)	r	r2	r2adj	Function
2.58 to 3.22	0.907	0.822	0.800	f(x) = −24.614 + 9.163x
3.22 to 4.03	0.954	0.910	0.899	f(x) = −12.182 + 8.925x
4.04 to 5.16	0.952	0.907	0.895	f(x) = −0.547 + 8.127x
5.16 to 6.92	0.923	0.853	0.834	f(x) = 4.817 + 8.320x
6.92 to 10.26	0.904	0.818	0.795	f(x) = 10.445 + 9.510x
10.26 to 20.01	0.887	0.787	0.761	f(x) = 13.327 + 11.834x

). The highest r²_adj with 0.899 was obtained for the surface area of the second-lowest mesopore classes, ranging from 3.22 to 4.03 nm. The same holds true for the correlation coefficient r with 0.954. Both r and r²_adj decrease with an increasing mesopore class. This indicates that the relationship between the adsorption capacity and the mesopores is stronger for mesopores with a smaller diameter. The assumption could be made that the correlation coefficient would be higher for micropores below 2.58 nm, as α-tocopheryl acetate might be smaller, assuming a size of 1.19 × 0.95 nm for vitamin E [25]. However, due to the weak correlation coefficients of the micropore volume (<2.58 nm) with the relative adsorption capacity of α-tocopheryl acetate (r = 0.228, r² = 0.052, and r²_adj = −0.066), we confirmed the findings of Hartmann et al.’s [34] study, who showed that the adsorption of vitamin E in mesoporous materials is larger than that in microporous adsorbents. One reason for this could be that adsorption into micropores is slower than that into mesopores. Slower or no loading of micropores could be due to molecules docking in channels to larger pores, and thereby blocking micropores [34]. The fact that adsorption mainly occurs in mesopores is advantageous because, according to Weber et al. [35], micropores are not accessible during animal digestion. This suggests that vitamin E is more likely to desorb from mesopores than from micropores.

3.4.3. Correlation of Adsorption Capacity and Biomass and Pyrolysis Processes

To test the hypothesis that the biomass and pyrolysis processes affect adsorption, the data were sorted. For the biomass, the main biomass was used, and biochar E, without specified biomass, was removed. However, as seven of the biochars contained mainly beech or had beech as the first biomass listed, and biochar I consisted of spruce and biochar B of poplar, no correlation could be tested. For correlation, biochars C and D were removed, due to unspecified pyrolysis processes. The remaining biochars were either produced with ‘PYREG’ (n = 6) or with a kiln (n = 2). Since no quantitative descriptors of the processes exist, the grouped adsorption capacities were compared using a t-test. As the normality failed (p = 0.008), a Mann–Whitney Rank Sum t-test was conducted. This test revealed no significant differences (p = 0.286). Thus, the variation between the biochars does not correlate with the pyrolysis process. Referring to the previously reported results and findings from the literature [46,58,63], it is assumed that the adsorption capacity depends on multiple internal and external factors.

3.5. Limitations of the Study

Our study indicates that the theoretical calculation of adsorption capacity from surface area and molecular size can be used as a tool to predict adsorption, select an adsorbent, and thus reduce, e.g., the number of trials or analyses. One limitation of this study is that we conducted single-point batch experiments, which is why no isotherm was determined for the adsorption of α-tocopheryl acetate into the pores of the different biochars. Among others, an isotherm could have been calculated using different concentrations of α-tocopheryl acetate, varying temperatures, and/or varying adsorption durations. We used a short adsorption time (10 min) and did not adjust the temperature or the concentration. For instance, Sarker et al. [25] varied the adsorption time from 6 to 24 h, whereas Chandrasekar et al. [49] varied the α-tocopherol concentrations from 0.25 to 40 g/L.

Another limitation of our study is that we did not adjust the moisture content of the biochars to the same level prior to the experimental determination of the adsorption capacity. The reason for this is that the moisture content varies depending on the biomass and production process [75], impacting its properties [64], such as the adsorption capacity. The aim of our study was to find the most suitable biochar to adsorb α-tocopheryl acetate, while considering biochars authorised as a feed additive for dairy cows. Thus, we used commercially available biochars, as purchased, to make the study as application-oriented as possible. With this, we also wanted to enable our results to be applied directly to, e.g., the feed industry.

4. Conclusions

Our hypothesis that an optimal adsorbent can be predicted by theoretical calculations using the surface area and the molecular size of the adsorptive was proven by the experimental adsorption capacity. The regression analysis revealed the strongest correlation, with r = 0.954, between the experimental adsorption capacity and pores ranging from 3.22 to 4.03 nm. Based on this, the number of studies to find an optimal adsorbent can be reduced. We propose that our findings could directly benefit the feed industry and can be transferred to other fat-soluble vitamins. Additionally, the hypothesis that the adsorption capacity directly depends on the biomass and production process was not directly supported. Our findings suggest that multiple internal and external factors influence the adsorption capacity.

Future studies should investigate, among others, the impact of pH modification and/or adjustment of moisture content on desorption properties. In this context, analysing the desorption from the pores of a high-adsorption-capacity biochar is highly relevant. On top of that, it would be interesting to investigate the theoretical and experimental adsorption capacity for, e.g., polar and hydrophilic molecules.