1 Introduction

Solid wastes are generated enormously around the globe because of the fast-paced rates of urbanization and industrialization. Several factors contribute to the concern over the proper management of these wastes, including a limited number of landfills, environmental challenges, and public awareness. People are consuming more and generating vast quantities of waste to achieve increasing living standards, while industries are booming to meet such demands. Consequently, these unsustainable patterns place pressure on the environment, which is already suffering, and serious action must be taken to reverse the damage and save the planet for future generations.

The use of thermochemical conversion techniques in converting waste to energy is gaining wide acceptance due to the many benefits resulting from the conversion of waste into energy, fuel, and other products [1,2,3]. In particular, biomass waste constitutes a source of energy that is both cheap and abundant [4]. Therefore, biomass conversion into energy requires a lot more attention. There are many routes that can be followed to achieve the benefits of waste to energy conversion. Pyrolysis, for instance, is a promising technique that needs more research in order to be commercialized and used on a larger scale [5]. This will enable concepts like sustainability and the circular economy to be employed and utilized more efficiently worldwide.

Biochar is a charcoal-like material that is full of benefits and is the result of the conversion of waste to products through the pyrolysis process [6]. Some of the applications of biochar include its use in agriculture in order to increase water retention and fertility of lands and soils, in addition to its use as a precursor for adsorbents like those used in water treatment [7]. The addition of biochar to the soil can reduce denitrification, thus resulting in less N2O emissions [8]; furthermore, several researchers have reported that adding biochar to the soil can boost crop yields by improving soil cation exchange capacity, water retention capacity, nutrient retention capability, improving soil fertility, and enhancing soil microbial activity, and pH [9,10,11]. Hence, it is important to understand the processes used to produce such biochars and other products. This can be achieved by the application of thermogravimetric analysis (TGA). The application of TGA involves studying changes in materials and products by following patterns related to change in weight loss with respect to temperature and time.

Globally, tomato is the second most used vegetable after potato, while carrot and cucumber are among the top ten most consumed vegetables. According to FAOSTAT, about 182 million tonnes of tomato, nearly 75 million tonnes of cucumber and gherkin, and almost 40 million tonnes of carrot and turnip are produced worldwide [12]. Due to the massive consumption of carrots, cucumbers, and tomatoes worldwide, huge quantities of these vegetable waste materials are produced every year. Despite having a high moisture content, the above vegetables have a high volatile content, which makes them potentially useful as sources of energy. Hence, in this study, carrot, cucumber, and tomato are considered, and their potential as feedstocks for pyrolysis is investigated. Many researchers have used TGA to investigate the pyrolytic conversion of biomass waste and food waste [13,14,15]. Specific examples of biomass waste and food waste included bamboo cane [16], camel manure [17], sewage sludge [18], and fruit waste [19]. The vast quantities of food waste generated are a strong motivation to use them in waste to energy conversion owing to their unique properties. This is especially true if methods like pyrolysis are used to transform them [20, 21].

The thermal characteristics and pyrolysis behavior of many vegetable wastes have been investigated by researchers. Table 1 presents excerpts of previous studies on pyrolysis characteristics of some top consumed vegetable wastes.

Table 1 Previous studies on pyrolysis behavior of vegetable wastes
Full size table

The table indicates that studies on thermal behavior and pyrolytic characteristics of some widely consumed vegetables such as tomato, cucumber, and carrot are yet to be made. Hence, in this study, the characteristics of pyrolytic decomposition of tomato, cucumber, and carrot wastes using TGA have been investigated. The study also investigates the pyrolytic response of the vegetable wastes’ binary (tomato + cucumber, tomato + carrot, and cucumber + carrot) and ternary (tomato + cucumber + carrot) blends which are quite novel. To produce value-added products from lignocellulosic biomass through pyrolysis, large amounts of feedstock would be required, which means blended biomass—rather than a single type of biomass—would be required. In addition, feedstock blending has a significant impact on the distribution of pyrolysis products. The pyrolytic degradation kinetics of all combinations (individual vegetable wastes, binary, and ternary vegetable blends) are determined employing two single-heating rate models such as Arrhenius and Coat-Redfern (CR). The information on the thermal characteristics and pyrolytic decomposition kinetics will be helpful in the design of pyrolyzer systems and in the calculation of the heat and mass balance of the process.

2 Materials and methodology

The three vegetable materials investigated in this study are carrots (Daucus carota subsp. Sativus), cucumbers (Cucumis sativus), and tomatoes (Solanum lycopersicum). These raw materials have been studied as a single-component system (carrot), binary component system (carrots + cucumbers), and the ternary component system (carrots + cucumbers + tomatoes). Prior to the TGA, the binary and ternary mixtures of components were prepared and mixed in an equal proportion by weight. This enabled the study of the different interactions between the components.

Initially, the raw materials (i.e., cucumbers, carrots, and tomatoes) were first dried using a hot-air oven at a temperature of 140 °C for 24 h. Next, the dried sample was chopped into smaller chunks, which was followed by grinding them into finer particles as required for TGA. The TGA was carried out from room temperature to 800 °C at two different heating rates, namely 5 and 10 °C/min under an N2 atmosphere (100 ml/min) employing the thermogravimetric analyzer (Discovery TGA 550, TA Instruments). As the purpose of the study is to produce biochar, the TGA experiments were performed at slow heating rates (5 and 10 °C/min). Additionally, slow heating results in effective heat transfer between particles, resulting in effective cracking/decomposition of particles. Approximately 7–18 mg of the raw materials was used for the thermogravimetric weight-change analysis. TGA was conducted by ramping the temperature up from room temperature to 100 °C using two different heating rates of 5 and 10 °C/min; keeping an isothermal condition at 100 °C for 10 min; ramping the temperature up to 800 °C using two different heating rates of 5 and 10 °C/min. Furthermore, the evolving gases were monitored using a mass spectrometer coupled to the thermogravimetric analyzer.

2.1 Materials characterisation

The materials analysis methods used to characterize these raw materials are:—proximate analysis on a wet and dry basis, for volatile content, fixed carbon, and ash content; elemental or ultimate analysis on a dry and ash-free basis for carbon, hydrogen, sulfur, nitrogen, and oxygen by difference. The standard followed for the determination of moisture content was a procedure by Choi et al. [30]. The instrument used for the proximate analysis was a SDT 650 (TA Instruments) Thermal analyzer, and the standard used for the proximate analysis was ASTM D7582–12. The elemental analysis of the feedstocks was carried out using a Euro-vector Euro EA 3000 CHN elemental analyzer. The ultimate analysis was conducted in line with standard ASTM D 3176–8.

2.2 Kinetic analysis models

For the purpose of pyrolyzer design, a detailed understanding of the pyrolysis-activation mechanism is essential. The TGA technique enables the study of the pyrolysis and activation behaviors of biomass materials. The formation of char from biomass consists of pyrolysis under an inert gas atmosphere to disintegrate the cross-linkage between carbon atoms. Although various kinetic models have been established for the study of pyrolysis mechanisms and kinetic parameters using TGA [31,32,33,34,35,36,37,38,39], in the present study, the Arrhenius and CR kinetic models were employed. The Arrhenius model is used because it is simple and assists in determining the order of reaction as such. In addition, the model had been used to determine the slow pyrolysis kinetics of lignocellulosic biomass such as sugarcane bagasse, coconut, coir pith, groundnut shell, casuarina leaves [40], sawdust, rice husk, and wheat husk, [41]. The CR model, like the Arrhenius model, does not require more than one heating rate to calculate kinetics and is used in this study because of its simplicity. The model was selected for its success in determining the slow pyrolysis kinetics of biomasses such as pine sawdust, olive waste, date palm trunk, seaweed [42], hemicellulose of Camellia Oleifera shell [43], and dry kitchen wastes [44].

2.2.1 Arrhenius model

The theoretical background of the model is presented below:

The kinetics of the devolatilization (pyrolysis) process can be given by

$$\frac{-dX}{dt}=k{X}^{n}$$
(1)

where \(X\) represents weight fraction undergoing decomposition, k corresponds to the reaction constant, and n is the order of the reaction.

The X and k can be presented as below:

$$X= \frac{w- {w}_{f}}{{w}_{0}-{w}_{f}}$$
(2)

where \({w}_{0}\) is the initial weight of the sample before pyrolysis decomposition, \(w\) is the weight of the sample at a particular time, while \({w}_{f}\) is the final weight of the sample after pyrolysis decomposition.

$$k=A{e}^{-RT}$$
(3)

where \(R\) is the universal gas constant (8.314 J mol−1 K−1) and \(T\) is the absolute temperature (K).

Substituting the values of X and k in Eq. (1) gives the below equation:

$$\mathrm{ln}\left[\frac{-1}{{w}_{0}- {w}_{f}} \frac{dw}{dt}\right]=\mathrm{ln}\left(A\right)-\left[\frac{E}{RT}\right]+n\mathrm{ln}[\frac{w- {w}_{f}}{{w}_{0}- {w}_{f}}]$$
(4)

where \(\frac{dw}{dt}\) is the rate of change in weight of the sample undergoing degradation?

Equation (4) takes the below form:

$$y=B+Cx+Dz$$
(5)

where \(y\) resembles \(\mathrm{ln}\left[\frac{-1}{{w}_{0}- {w}_{f}} \frac{dw}{dt}\right]\), \(x\) resembles \(\frac{1}{T}\), and \(z\) resembles \(\mathrm{ln}[\frac{w- {w}_{f}}{{w}_{0}- {w}_{f}}]\).

While the constant \(B\) and coefficients of \(C\) and \(D\) correspond to

$$B=\mathrm{ln}\left(A\right), C=[\frac{-E}{R}] , D=n$$
(6)

2.2.2 Coats-Redfern model (CR)

In general, the CR model is given by

$$\mathrm{ln}\left[\frac{g\left(\propto \right)}{{T}^{2}}\right]=\mathrm{ln}\left[\frac{AR}{\beta E}\right]-\frac{E}{RT}$$
(7)

While the reaction mechanisms of the CR model can be broadly classified into (i) based on the rate-determining reactions; (ii) based on reactions influenced by the diffusion of solids; (iii) based on the interphase transfer of products [44]. A few most common reaction equations and the functions of \(g (\propto )\) in the CR model are presented in Table 2.

Table 2 Reaction models of the CR model
Full size table

2.3 TGA-MS analysis

The TGA-MS analysis provides information on the thermal behavior of feedstocks as well as details on the distribution of products. A mass spectrometer (MS) makes use of differences in the mass-to-charge ratio of ionized molecules to identify species or components, and those mass-to-charge ratios are referred from the National Institute of Standards and Technology (NIST) database. Since it is unviable to detect all gas components in the syngas, only the H2, CH4, H2O, and CO2 gas components were analyzed, considering the below equation.

$$\mathrm{Biomass waste}\stackrel{\Delta }{\to }{{\mathrm{Char}}_{(s)}+ {\mathrm{Tar}}_{(l)}+ {\mathrm{H}}_{2}{\mathrm{O}}_{(g)}+ \mathrm{H}}_{2(g)}+ {\mathrm{CO}}_{(g)}+ {\mathrm{CO}}_{2\left(g\right)}+ {\mathrm{CH}}_{4\left(g\right)}$$
(17)

The gas components H2, CH4, H2O, and CO2 were analyzed, referring to atomic mass units (amu) 2 (H2), 15, 16 (CH4), 17, 18 (H2O), and 44 (CO2), respectively. The mass spectra of N2, C2H4, and C2H6 have the same amu 28. Since the TG-MS analysis was conducted under an N2 atmosphere, the data for C2H4 and C2H6 was not presented as N2 would have disturbed the amu 28 [45].

2.3.1 Semi-quantitative gas analysis

Radojevi´c et al. [45] proposed a new method for estimating the quantity of gas components present in syngas. Based on the method, the volumetric share per unit mass of the gas component is determined using the below equation.

$${v}_{i}\left[\frac{{m}^{3}}{kg}\right]= \frac{1}{\beta } \frac{{v}_{c}}{m} \underset{T1}{\overset{T2}{\int }}\frac{{IC}_{i}}{{IC}_{c}} dT$$
(18)

where \({v}_{i}\) represents the volumetric share of gas component (m3 kg−1), \(\beta\) refers to the heating rate of pyrolysis process (K/min), \({v}_{c}\) corresponds to the volumetric flow rate of carrier gas (ml/min), \(m\) denotes the mass of the feedstock sample used for the analysis (mg), \({IC}_{i}\) refers to the peak ion current of the gas component (A), \({IC}_{i}\) represents the ion current of the carrier gas (A) (usually the carrier gas has a constant ion current value), \(T1\) and \(T2\) are the lower and upper limits of the integral function, indicating the temperature range used in the calculation. The term \(T1\) is the logarithm of ion current of gas component while \(T2\) is the temperature that corresponds to the peak ion current of the gas component.

3 Results and discussion

3.1 Materials characterization

Any study of biomass foodstuffs needs to describe the characteristics and properties of the raw materials. When it comes to food waste, an elemental analysis on a dry basis as well as the moisture content is needed. Moisture content becomes critical in thermal treatment studies because it can range from a few percent to over ninety percent, consuming over sixty percent of the heat energy in the thermal processes. The classical analyses for the three vegetables under consideration, namely, carrots, cucumbers, and tomatoes, are now presented.

3.1.1 Proximate analysis

Due to the extremely wide range of moisture contents depending on the source and processing methods utilized, the proximate analyses are presented on a dry basis. The moisture contents of the natural vegetables are 90.01 wt. %, 96.07 wt. %, and 94.72 wt. % for carrots, cucumbers, and tomatoes, respectively. Experiments were performed at least in triplicate to obtain three consistent values that were within ± 5%. The results of the proximate analysis of the three vegetable food wastes are presented in Table 3.

Table 3 Proximate analysis of the chosen vegetables (this study)
Full size table

3.1.2 Ultimate analysis

The results of the ultimate analysis of the three vegetables are presented in Table 4. The values are based on the average of three measured results for each sample falling within ± 6%. The ultimate analysis results of these vegetables reported elsewhere are also presented in the table.

Table 4 Elemental analysis results for the food samples dried for 24 h at 140°C, using CHNS reactor
Full size table

3.1.3 Heating value

The higher heating values (HHV) of the foodstuffs is calculated based on the results of the elemental analysis using the correlation proposed by Channiwala and Parikh [49] which is given below:

$$HHV \left(\frac{kJ}{kg}\right)=349.1C+1178.3H+100.5S+103.4O-15.1N-21.1 Ash$$
(19)

The above correlation is used since the composition of all elements are well within the allowable limits suggested by the researchers, i.e., C: 0–92%, H: 0.43–25%, N: 0–5.6%, O: 0–50%, Ash: 0–71%, HHV: 4745–55,345 kJ/kg. The obtained heating values of the vegetable wastes are compared with the heating values of other vegetable wastes (cauliflower [23], lettuce [23], coriander [23], banana flower [29], pea [28], potato peel [22]) (see Fig. 1).

Fig. 1
figure 1

Heating values of vegetable food wastes

Full size image

It can be noted that the cucumber and tomato wastes have high heating values. The high heating values of the vegetable food wastes are due to their high carbon content and low ash and oxygen content. On the other hand, carrot waste exhibits a low heating value. The low heating value of the waste is because of its low carbon and high oxygen content.

3.2 TGA profiles and biochar yields

The TGA profiles at the two heating rates, 5 and 10°C/min, are shown in Figs. 2, 3, and 4. The three profiles are for carrots only carrots and cucumbers binary blend and for carrots, cucumber, and tomatoes ternary blend in Figs. 2, 3, and 4, respectively. The TGA data are represented by the dashed lines. The differential thermal gravimetric, DTG, percentage mass losses are shown by the dotted lines and the percentage values on the left-hand scale. The mass quantities at each temperature can be abstracted from the TGA versus temperature curves and are presented in Tables 5 and 6 for the two heating rates at 5 and 10 °C/min.

Fig. 2
figure 2

TGA/DTGA profiles for the carrots single component at 5 and 10°C/min heating rates

Full size image
Fig. 3
figure 3

TGA/DTGA profiles for the carrots and cucumbers blend at 5 and 10°C/min heating rates

Full size image
Fig. 4
figure 4

TGA/DTGA profiles for the carrots, cucumbers, and tomatoes ternary blend at 5 and 10°C/min heating rates

Full size image
Table 5 Dry mass yields, as mass fractions, while taking the moisture content into account and dividing by the dryness fraction for 5°C/min
Full size table
Table 6 Dry mass yields, as mass fractions, while taking the moisture content into account and dividing by the dryness fraction for 10°C/min
Full size table

Table 5 presents the mass yields at different temperatures on a dry mass basis by taking the moisture content into account and dividing by the dryness fraction for the mass values 5 °C/min. For a system at a fixed heating rate, the height of the first peak of the DTGA corresponds to moisture content (wt.%), which is then converted to the solid dryness fraction (100—moisture content) and then used to divide the original mass yields to arrive at the corrected dry mass yields in Tables 5 and 6. The intention of these tables is to see if blending vegetable food wastes made any difference to the biochar yields.

The data in Table 5 show a steady decreasing trend in the mass fraction biochar product yield with increasing temperature for all three systems studied. The mass fractions showed, when multiplied by 100, represent the solid percentage yields. The single-component carrot has a higher yield at each temperature. This could be caused by the other two systems (binary and ternary) having moister and softer interiors associated with higher cellulose contents than carrots.

In Table 6, for the single component, the char yield at 800 °C is 0.457, the char yield for the binary is 0.415, and for the ternary component, the yield is 0.383. At 700 °C, the yields are 0.468, 0.436, and 0.406 for single, binary, and ternary systems, respectively. There is a slight trend showing a decrease in the mass yield with blending at both 700 and 800 °C, but at the heating rate of 5 °C/min, the differences in the yields are quite small. Considering the data at the higher heating rate of 10 °C/min and 800 °C, the single component carrot has the highest yield while the ternary component has the lowest yield.

As in the case of the lower heating rate, there is a slight trend in decreasing biochar formation with an increasing number of blended materials. In addition, the biochar yield is lower at the more rapid heating rate, which is well established in pyrolysis studies, so concluding there is more syngas and bio-oil produced with increased heating rate.

Figures 5 and 6 present the percentage char yields at different temperatures based on dry solids content at 5 and 10 °C/min, respectively.

Fig. 5
figure 5

Percentage char yields at different temperatures based on dry solids content at 5°C/min heating rate

Full size image
Fig. 6
figure 6

Percentage char yields at different temperatures based on dry solids content at 10°C/min heating rate

Full size image

Figures 5 and 6 demonstrate a slight difference in biochar yields at 600, 700, and 800 °C. The char yield for the single component is minimal at both heating rates, whereas it is high for the ternary system, as seen in the figures. The char yield of the binary component lies between the char yield of single and ternary components. In the case of 5 °C/min, the biochar yield of the single, binary, and ternary component was observed to be 44.2%, 43.0%, and 42.2% respectively, whereas, in the case of 10 °C/min, the biochar yield of the single, binary, and ternary component was observed to be 45.7%, 41.5%, and 38.3%, respectively. The biochar yields of the current work are compared with biochar yields of other similar works in Table 7.

Table 7 Literature values for biochar yield at high temperatures
Full size table

3.3 Food waste decomposition mechanism

Inspection of Figs. 2, 3, and 4 at both heating rates, the total mass losses at 800 °C temperature is approximately 58–64%; however, the main volatilization fraction of the food waste biomasses is taking place between 180 and 510 °C. This significant volatile release may be assigned to the thermal decomposition of the three major food biomass constituents, namely: lignin, cellulose, and hemicellulose. These three constituent compounds have been well established as being responsible for this major mass loss during the pyrolysis process [51].

Furthermore, the presence of specific peaks can be observed in the DTGA results curves, representing the maximum reaction rates at both the heating rates. The initial peak can be seen in the region of 100–120 °C and is attributed to the evaporation of the moisture, for the current samples (because of the drying pretreatment), the mass loss is low in the region of 3–7%. The second DTGA peak can be observed in the region 250–270 °C for both heating rates 5 and 10 °C/min. This peak is principally attributed to the pyrolytic decomposition of hemicellulose. Several other studies have reported in the scientific literature that hemicelluloses decompose in the 150–300 °C temperature range [52,53,54,55]. The third and final main peak is observed at 300–320 °C for both the heating rates. This peak is mainly attributed to cellulose itself, which generally undergoes degradation at quite high temperatures i.e. 300–400 °C [31, 52, 54]. In addition, the first biomass constituent to undergo decomposition at small temperatures is lignin itself, and its degradation occurs slowly at a very low reaction rate and proceeds right through the pyrolysis period until 800 °C with some peaking around 500 °C. Consequently, this implies that small sections of the second and third peaks on the DTGA curves are associated with the decomposition of lignin [56]. Nevertheless, there is quite a degree of similarity between curves generated at the two different heating rates applied in this study. The peaks themselves represent the same structure and composition, and the mass loss trends follow the same direction, particularly at the higher temperatures, but the quantitative amounts, although the differences are quite small, do represent a significant change in the relative quantities of product formation. This trend in product distribution change due to variations in the heating rate has also been reported in the literature [53, 57]. In the present study, it should be noted that the low heating rates were selected to investigate the slow pyrolysis, which has been shown to produce higher yields of product biochar.

3.4 Kinetic analysis

3.4.1 Arrhenius model

The TGA data are used to define Eq. (4), and the triplet kinetic parameters-pre-exponential factor, activation energy, and order of reaction were determined using a multiple linear regression equation in Microsoft excel. Only the kinetic parameters for the active pyrolysis zone (140–620 °C) were determined. The determined kinetic parameters for the food wastes at two different heating rates are presented in Table 8. The obtained R2 values lie between 0.80 and 0.98, indicating the good fitness of the model. The R2 values for all the runs were in the range of 0.80 and 0.98.

Table 8 Kinetic parameters determined using Arrhenius model
Full size table

As can be seen from the table that the pre-exponential factor values for all the runs are in the range of 7.54 × 104–8.62 × 106 (s−1). This low range of (< 109 s−1) pre-exponential factor values indicates that the pyrolysis of the food wastes could be a surface reaction or closed complex reaction [58]. The activation energy values for the samples were between 15 and 26 kJ/mol. This low range of activation energy values signifies that the pyrolysis of the wastes would be simple and smooth. The reason for these low activation energy values could be due to their high cellulose and hemicellulose content and low lignin content. In general, the pyrolysis of cellulose and hemicellulose is simple and hence requires low activation energies while the pyrolysis of lignin is intricate, hence it needs high activation energies [40]. The order of reaction for all the wastes was between 0.75 and 0.89.

The obtained values of the kinetic parameters are compared with the kinetic values of other works. Mansaray and Ghaly [59] were one of the pioneers to employ the Arrhenius model for determining the pyrolytic kinetics of plant biomass wastes. The researchers used the model to determine pyrolysis kinetics of different varieties of rice husk. The estimated pre-exponential factor values were in the range of 1.97 × 1012–2.03 × 1015 (s–1), while the activation energy values were between 142 and 189 (kJ mol–1), and the orders of the reaction were between 0.70 and 0.83. Kumar et al. [60] employed the model to determine the kinetic parameters of corn stover pyrolysis. The triplet kinetic parameters for the active zone were in the range of 1.35 × 104–7.33 × 104 (s−1), 57–63 (kJ mol−1), and 0.74–0.73, respectively. Wu et al. [61] used the model to estimate the kinetic parameters of dairy manure pyrolysis. The pre-exponential factor, activation energy, and order of reaction for the active pyrolysis region were estimated to be 7.03 × 106 (min−1), 93.63 (kJ mol−1), and 6.37, respectively.

The kinetic parameters are primarily influenced by the nature of feedstocks, the composition of components present in the feedstocks, temperature range, and heating rate. Hence, a difference in the values of kinetic parameters is observed in all the studies.

3.4.2 CR model

In this model, plotting \(\mathrm{ln}\left[\frac{g\left(\propto \right)}{{T}^{2}}\right]\) vs \(\frac{1}{T}\) gives a slope from which activation energy (E) is calculated. Among the reaction models, only the models that are good fitting with the experimental values are considered, and the kinetic parameters obtained from them are provided in Table 9.

Table 9 Kinetic parameters of vegetable food waste samples determined using different reaction models
Full size table

As can be seen that first-order reaction (F1) is the best-fitting model. This indicates that first-order reaction is dominant in the active pyrolysis zones of all chosen vegetable food wastes. The pre-exponential factor values for all the wastes were below 102 (s−1), indicating that the pyrolysis of these wastes would undergo degradation smoothly without any complication. The obtained activation energy values ranged between 28 and 77 kJ mol−1. The F1 mechanism exhibited the lowest E-value, while the D3 model demonstrated the highest E-value.

3.5 Thermodynamic properties

The kinetic parameters of the good fitting models were used to determine the thermodynamic properties such as enthalpy (ΔH), Gibbs free energy (ΔG), and entropy (ΔS). The thermodynamic properties values were calculated using the following equations:

$$\Delta H=E-R{T}_{p}$$
(20)
$$\Delta G=E+R{T}_{p} \mathrm{ln}(\frac{{k}_{b}{T}_{p}}{hA})$$
(21)
$$\Delta S=\frac{\Delta H-\Delta G}{{T}_{p}}$$
(22)

where \(R\) represents the gas constant having the value 8.314 J mol−1 K−1, \({T}_{p}\) corresponds to the peak decomposition temperature (K), \({k}_{b}\) denotes the Boltzmann constant (1.38 × 10−23 J K−1), and \(h\) stands for Planck constant (6.63 × 10−34 Js). The thermodynamic properties of the vegetable food wastes determined using different reaction mechanisms of the CR model are listed in Table 10.

Table 10 Thermodynamic parameters of the vegetable food waste samples calculated using the CR model
Full size table

\(\Delta H\) is a state function of thermodynamic phenomena which indicates whether a system takes or emits heat. As can be noted from the table that \(\Delta H\) values are between 26 and 72 kJ mol−1. The positive values of \(\Delta H\) for all the models signify that the energy is required from an external source to reach the transition state. This indicates that only small external energy is sufficient to pyrolyze these vegetables’ food wastes. The \(\Delta G\) values are positive and are consistent (162–202 kJ mol−1). The positive values signify that the degradation phenomenon is an endergonic reaction. Negative \(\Delta S\) values are observed for all the samples signifying the creation of activated complexes while small \(\Delta S\) values indicate the slow reactivity of the samples.

3.6 Syngas analysis results

3.6.1 TGA-MS profiles

The TGA-MS profiles of carrot, carrot–cucumber, and carrot–cucumber–tomato are presented in Fig. 7.

Fig. 7
figure 7

TGA-MS profiles: a carrot at 5 °C/min, b carrot at 10 °C/min, c carrot–cucumber at 5 °C/min, d carrot–cucumber at 10 °C/min, e carrot–cucumber–tomato at 5 °C/min, and f carrot–cucumber–tomato at 10 °C/min

Full size image

As can be seen from Fig. 7, the peak release of H2O and CO2 occurs relatively at the same time, while the peak release of CH4 takes place a little later, followed by the evolution of H2. The release order of gases took place in the following order H2O, CO2, CH4, and H2, which is consistent with the observation of Ma et al. [62]. In general, the liberation of the above gases is a complicated phenomenon as some gases evolve due to thermal degradation of feedstocks, some gases are adsorbed ones while the rest are released owing to secondary reactions among the product gases [63, 64]. It is significant to note that the ion-current intensities of H2O and CO2 are strong, indicating a high presence of hydroxyl groups and oxygen atoms in the selected food waste samples.

3.6.2 Pyrolysis stages

In general, based on the decomposition trend of the samples, the TG-DTG curves are classified into four stages: zeroth stage (˂150 °C), the first stage (150–250 °C), the second stage (250–500 °C), and the third stage (> 500 °C). The zeroth stage corresponds only to the dehydration phase, wherein no gas products are evolved, and it can be assumed that no pyrolysis occurs in this stage. In the first stage, gases such as CH4, C2H4, C2H6, and H2O are liberated. No permanent gases such as H2, CO, CO2, and O2 are produced in this stage. In the second stage, most gas products except H2 are generated. The pyrolysis of biomass primarily occurs in this stage. In the third stage, biomass pyrolysis occurs with the production of CO2 and H2. The interaction between CO2 and H2 also takes place in this stage. Table 11 presents the detected gas components and their respective pyrolysis stages. The table also presents the details of peak ion current and peak temperature of the gas components, which are very vital for the quantification of the gas components.

Table 11 Evolved gas components and their respective pyrolysis stages
Full size table

From Table 11, it can be noted that the temperature of H2O peak lies between 202 and 332 °C, the temperature of CO2 peak lies in the range 268–329 °C, the temperature of CH4 peak is ranges between 267 and 524 °C, and the temperature of H2 peak falls in the 569–654 °C temperature range. The observed temperature ranges indicate that the release of H2O falls under the first and second stage of pyrolysis, the evolution of CO2 falling under the second stage of pyrolysis, the liberation of CH4 belongs to the second and third stage of pyrolysis, while the discharge of H2 belonging to the third stage of pyrolysis.

3.6.3 Effect of heating rate on the syngas components yield

Based on the equation, the volumetric compositions of H2, CH4, H2O, and CO2 components were calculated, and the effect of heating rate on the syngas components yield is illustrated in Fig. 8.

Fig. 8
figure 8

Effect of heating rate on the syngas component yield

Full size image

It can be seen from the figure that increasing heating rate from 5 to 10 °C/min produced an insignificant change in the H2 yield of all feedstock samples. However, an elevation in heating rate brought about a decrease in the yields of CH4. An increase in heating rate produced a decrease in H2O yields of single and binary samples, while in the ternary sample an increase in H2O yield is noted. An elevation in heating rate also produced a mixed trend in the yield of CO2, i.e., a decrease in CO2 yield in the case of single and ternary samples and an increase in CO2 yield in the case of the binary sample.

3.6.4 Effect of blending on the syngas component yield

The effect of blending on the syngas components yield at heating rates 5 and 10 °C/min is presented in Fig. 9.

Fig. 9
figure 9

Effect of blending on the syngas components yield at heating rates: a 5 °C/min; b10 °C/min

Full size image

It is interesting to note that at both heating rates, the H2 yield is the highest for the ternary mixture followed by the single feedstock sample and then by the binary mixture. This indicates that the ternary mixture combination is encouraging the H2 yield. At heating rate of 5 °C/min, the CH4, H2O, and CO2 yields of a single feedstock sample are more than the binary and ternary mixtures. At a heating rate of 10 °C/min as well, apart from the H2O yield, the yields of CH4 and CO2 are more than the binary and ternary mixtures. This indicates that the carrot favors more CO2 and CH4 production than cucumber and tomato upon pyrolysis. It is evident from the figure that the syngas yields of binary and ternary mixtures differ considerably and can be attributed to the differences in the cellulosic and hemicellulosic contents in the binary and ternary mixtures. Different syngas compositions in binary and ternary mixtures signify that the interaction between carrot–cucumber and carrot–cucumber–tomato is different. It also indicates the synergistic effect of feedstock samples.

4 Conclusion

The current work focuses on the understanding of the pyrolysis process and the generation of biochars from food waste. According to the TGA results, the endothermal peaks of hemicellulose, cellulose, and lignocellulose degradation took place at ranges close to those reported in the literature studies, namely, 220, 315, and 505 °C, respectively. The biochar yields between 200 and 800 °C were in the range 82.7 to 40.8 w/w% and 82.3 to 38.3 w/w% at 5 and 10 °C/min, respectively. The biochar yields for the food waste blends were slightly lower than the yields for the individual food type. The gas analysis indicated that the ternary mixture produced the highest H2 yield, followed by the single feedstock sample and then by the binary mixture. Different syngas compositions in binary and ternary mixtures signify that the interaction between carrot–cucumber and carrot–cucumber–tomato is different. Further research should focus on the product distribution and yields from blends of food wastes but also on blends of food wastes with other types of waste to study these synergistic effects.