A master plot approach on rice husk decomposition through pyrolysis and kinetic elucidation using iso-conversional models

Abstract

The present study aimed at predicting the intricate mechanism followed by the pyrolysis of locally available rice husk, uses noval modified master plots through continuous slope-based differential techniques. An average apparent activation energy was found in the range of 187.29 (KAS model) and 199.85 kJ mol⁻¹ (OFW model), respectively. The rice husk revealed higher volatile matter; 54.51 wt-%, HHV; 18.42 MJ kg⁻¹and a critical pyrolysis zone around 250 to 550 °C with two sharp peaks through differential thermogravimetric analysis. Using Karl–Pearson correlation, experimental curve showed highest correlation coefficient of 0.94 and 0.75 with respect to A2 and A3 models (Avrami–Erofeev) for g(α). However, for Z(α), the process is limited by first-order multi-diffusion processes (D1) coupled with solid phase surface reaction (F3) showing correlation coefficient of 0.8. The experimental data worked out the generalized model of f(α) as f(α) = 12.74α³ – 24α²+ 12.6α - 0.92 with R² value of 0.99, narrating the usage of master plots for better understanding of the process.

Keywords

activation energy master plots pyrolysis rice husk

Introduction

The demand for energy sources is increasing at an alarming pace. For instance, numerous countries have pledged to set net zero emissions by 2050. However, the challenge looming in the forefront is getting stable, clean, and affordable alternative energy sources, as depicted by goal 7 of 17 sustainable development goals (SDGs).¹ One promising avenue is the harnessing of bioenergy from biomass. Biomass energy, also known as bioenergy, is a renewable energy source with significant potential to cater to energy demand and its transition towards low carbon economy.²

Biomass comprises of organic materials derived from plants, animals, and micro-organisms. It includes agricultural leftovers, kitchen residue, biomedical waste, and plant remains. India is one of the leading producers of agri-based biomasses, with an estimated production of 750 million metric tons (MMT) annually. However, only a meager proportion of 30% is currently utilized for bioenergy conversion.³ To capitalize on the bioenergy potential, one must also invest attention in other sources of Agri-based feedstock. Rice husk could be a potential candidate by virtue of its abundance in production and its chemical association.⁴ India is accountable for a staggering 20% of the world's rice production, making it the second-largest global producer with an annual output of 130 MMT.⁵ This is an awe-inspiring accomplishment that cannot be overlooked. Rice comprises 20% to 22% rice husk and is generated as one of the by-products in the rice milling plants, thus amounting to a production of about 28.6 MMT annually.⁶ Around 40% cellulose, 30% lignin group, and 20% silica are contained in rice husk.⁷ Rice is an essential food for the people of Kashmir Valley and plays a crucial role in the economy of Jammu and Kashmir, India. This state covers almost 38% of the total gross cropped area, and paddy is the most important cereal crop. The annual rice production in Jammu and Kashmir is about 0.58 MMT, with major districts like Anantnag, Jammu, Baramulla, and Pulwama contributing significantly to this production.

Interestingly, since rice husk makes up around 20% to 22% of rice, Jammu and Kashmir produces around 0.12 MMT, approximately.^5,8 Its combustion in the open fields increases carbon dioxide content and other environmental pollutants. Animal feed, energy resources (biofuel), organic fertilizer additive, making bricks, and production of bioethanol, bio-char, and composites are the various applications of rice husk instead. Paddy species, rice variety, fertilizer type, soil chemistry, climate, and geographical location differences are reasons rice husk composition varies from one sample to another.⁹

The conversion techniques of biomass to bioenergy include biochemical and thermochemical conversion. It has been reported that thermochemical conversion is relatively more efficient in dealing with biomass pyrolysis and, therefore, is used by many researchers.¹⁰ The thermochemical conversion process includes pyrolysis, incineration, gasification, and torrefication. The main principle by which the conversion process proceeds is the elimination of undesirable intermediates, including process optimization.¹¹ Pyrolysis has resulted in being one of the emerging solutions for bioenergy conversion of biomass. The major products obtained are char, bio-oil, and gas. The process is an endothermic reaction in which the decomposition of biomass takes place at higher temperatures. The process leads to the breakdown of polymeric structures into lower organic compounds.¹⁰ Recent years have observed a surge in studies primarily focusing on bioenergy conversion through pyrolysis of different biomasses.¹² Recent studies have unveiled the immense potential of various biomass feedstock when subjected to pyrolysis. Among other biomasses studied, the rice husk has shown a promising potential, as portrayed by several studies.^13,14

The latest reports on the subject include the study of thermokinetics of coconut husk, fast pyrolysis of rice husk, pyrolysis of rice husk using iso-conversional modeling, and catalytic pyrolysis of rice husk.^15–17 The studies clarified that the optimum temperature for the process is 400 to 600 °C. However, the findings were circumscribed in the field of determination of optimum pyrolytic temperatures and application of possible products, including char and bio-oil. Earlier, the pyrolysis of garlic husk using iso-conversional modelling was performed, which showed the apparent activation energy values of 153 to 155 kJ mol⁻¹.¹⁸ One more study conducted rice husk pyrolysis and revealed the optimum pyrolytic temperature of 427 °C along with the bio-oil yield of about 35.5 wt-%.¹⁹ Significant research has set a benchmark over the years to optimize the process parameters along with enhancement of yield and product quality of pyrolysis of rice husk. Over the past few years, several innovative approaches have been developed to enhance the efficiency and yield of rice husk (RH) pyrolysis.^20,21 Various techniques are used to study the process parameters, including the mechanism of complex pyrolytic reactions of different feedstocks. Recently, various biomasses were studied by combining iso-conversional models with MATLAB, Aspen, and computational fluid dynamics (CFD) models.²² The results were promising, with mild limitations of overfitting data. The model's applicability depends upon the number of parameters like feed type, reactor type, and degree of freedom of the system.²³ The researchers are currently studying the master plots model-free method, and its findings have revealed promising results. Recently, the same technique was used to analyze the mechanism of sewage sludge, polypropylene and high-density polyethylene.²⁴ The empirical function for switch grass pyrolysis was developed using master plot techniques.²⁵ Another study reported thermokinetic analysis of soyabean stalk, employing master plots for mechanism determination. The study efficaciously summarized the findings that different reaction mechanisms governed degradation in different conversion ranges.²⁶ Thus, the in-depth review of literature summarizes the urgency of evaluation of bioenergy potential of rice husk and affirms the role of iso-conversional models in association with master plots for the mechanism determination.

In our study, we developed an experimental strategy for studying rice husk thermokinetics using pyrolysis by applying iso-conversional models in association with master plots for mechanism determination and its kinetic analysis. We theorized that the breakdown of rice husk occurs through a multi-step process because of its intricate chemical composition. We expected the apparent activation energy of this breakdown to vary with temperature due to changes in reaction routes and energy barriers. We suggested that iso-conversional models in tandem with master plots might well describe the mechanism and kinetics of rice husk pyrolysis at various heating rates, allowing for a thorough comprehension of the process. Our research intended to confirm hypotheses and enhance the understanding of rice husk pyrolysis dynamics, having implications for both fundamental science and practical applications in biomass utilization.

The study embarks on the novel technique by employing the modified master plot method (based on a continuous slope-based technique) to determine the rate kinetics and reaction mechanism of locally available rice husk feedstock. The study aims to enhance the master plot algorithm's precision by truncating all possible errors that arise during integral approximations. In addition to the study of bioenergy potential of rice husk the objectives also includes determination of the process parameters of rice husk pyrolysis, kinetics and the possible mechanism of solid-state complex pyrolytic reaction. Furthermore, statistical analysis was developed and validated the empirical function of RH pyrolysis.

Materials and methods

Sample pre-treatment and experiment design

The biomass feedstock was extracted from three different rice varieties (Mushk budij, Jehlum 448, and K-332), which were collected from three different locations in the Kashmir valley, India. The three varieties were blended in equal proportion by weight (1:1:1) and ground to a fine powder to a uniform size of around 100 to 200 mesh. The product was then washed with distilled water to remove impurities and filtered. Finally, the sample was sun-dried for 48 hours followed by oven drying at moderate temperature (60–65 °C). The prepared samples of RH were stored carefully for further experimental analysis. The experimental design starts with inspecting physicochemical viability by performing the tests of proximate and ultimate analysis of the feedstock. Then, the random sample of feedstock is tested for its thermal degradation characteristics using TGA. The data generated is given its final form by removing the data containing possible errors. The final data is fitted using various models and model-free methods to evaluate reaction kinetics and their mechanisms.

Physicochemical characterization

In order to assess the viability of biomass waste as a fuel source, it is imperative to conduct a proximate and ultimate analysis. It is crucial to adhere to the standard procedures prescribed by the ASTM, such as (E871-82 through E872-82, E1534-93, 2013), for determining the parameters. The higher heating value (HHV, MJ kg⁻¹) of the RH samples was determined both theoretically and experimentally. The values from proximate analysis were substituted in the correlation²⁷ to calculate theoretical value.

HVV = 19.2880 – 0.2135 \frac{VM}{FC} - 1.9584 \frac{Ash}{VM} + 0.0234 \frac{FC}{Ash}

(1)

Where VM and FC are volatile matter (wt-%) and fixed carbon (wt-%) calculated through proximate analysis. Experimentally, the HHV was determined in a digital Bomb Calorimeter (Advance, IP- 3) following a standard ASTM code.²⁸ The prepared samples as discussed above were placed in the crucible of the bomb. The bomb was closed and pure oxygen (99.9%) was supplied at high pressure of 5 bar. The samples were tested thrice for checking erroneous results.

Thermal degradation analysis using TGA/DTG

The thermal degradation of biomass feedstock was studied through TG (thermogravimetric) and DTG (differential thermogravimetric) analyses. The thermal decomposition of the biomass sample was conducted under non-isothermal conditions using the Exstar TG/DTA 6300 system within the temperature range of 25 to 1000 °C. Several runs were carried out at three different heating rates of 20, 50, and 100 °C min⁻¹ in a nitrogen atmosphere at a flow rate of 200 mL min⁻¹. A sample of mass 10.37 mg was used in each run, and the accuracy of the data was maintained by repeating the experiments at least three times at each heating rate. The TG, DTG, and other related curves were obtained and analyzed to study the kinetic and reaction mechanism parameters using two iso-conversional methods along with a novel master plot method.

Kinetic interpretation and mechanism determination

In order to investigate the rate kinetics of pyrolysis reactions from TGA/DTG, various models are being employed in contemporary times.²⁹ In this study, two iso-conversional kinetics models are being used, namely, the Kissinger–Akahira–Sunose (KAS) and Ozawa–Flynn–Wall (OFW) model and the reaction mechanism is determined by the model-free coupling of Coats–Redfirn (CR) with Master plot method of mathematical modelling. The modelling is based on few fundamental assumptions, like solid-state reaction model, non-isothermal heat treatment and negligible heat and mass transfer resistances.^29,30

This one-step global mechanism of solid-state reaction sheds light on the chemical reactions involved in the process. The mass balance over the system leads us to the fundamental equation of pyrolysis.

Biomass \Rightarrow volatiles + Biochar + Bio oil

The general equation for the thermal decomposition of solid biomass is given as

\frac{d α}{d t} = k (T) f (α)

(2)

Where k is velocity constant.

One possible way to represent the heterogeneous function of uniform kinetics reactions, f(α), is by utilizing the following expression.

f (α) = (1 - α)^{n}

(3)

n signifies the order of reaction and

α

is the conversion, which is represented as;

α = \frac{m_{o} - m_{i}}{m_{o} - m_{f}}

(4)

Where m_o is initial weight of the sample, m_f is the final weight of the sample and m_i is the sample weight at time t.

The Arrhenius equation generally represents the function k(T) as,

k (T) = A_{o} \exp (\frac{- E}{R T})

(5)

The variable E is the apparent activation energy (kJ mo1⁻¹), and R is the universal gas constant (8.314 J mol⁻¹ K⁻¹). A_o is the pre-exponential factor (min⁻¹), and T is the absolute temperature (K). From equations (2) and (5), a mathematical rearrangement yield,

\frac{d α}{d t} = A_{o} \exp (\frac{- E}{R T}) f (α)

(6)

Introducing the heating rate

β = \partial T / \partial t

. Equation (6) becomes

\frac{d α}{d T} = (\frac{A_{o}}{β}) \exp (\frac{- E}{R T}) f (α)

(7)

Equation (7) can be arranged and integrated using initial conditions of α = 0 at T = T_o yielding

g (α) = \int_{0}^{α} \frac{\partial α}{f (α)} = \int_{T_{o}}^{T} \frac{A_{o}}{β} \exp (- \frac{E}{R T}) \partial T

(8)

Further mathematical rearrangements yield

g (α) = \frac{A_{o} E}{β T} \int_{x}^{α} (- u) d u = \frac{A_{o} E}{β R} Q (x)

(9)

where x = E/RT, the term Q(x) on RHS of (9) has no exact solution. Thus, equation (9) needs to be solved by approximations. The models based on iso-conversional methods are dependent on the approximations. The two iso-conversional kinetic models then yielded KAS and OFW models.

Kissinger–Akahira–Sunrose (KAS) and Ozawa–Flynn–Wall (OFW) model

The KAS model has been widely used for the kinetic study of biomass pyrolysis.³¹ In the KAS model, mathematical approximations for the exponential term is assumed, and after approximation and rearrangement, the solution is given as,

\ln (\frac{β}{T^{2}}) = \ln (\frac{A_{o} E}{R g (α)}) - \frac{E}{R T}

(10)

While plotting ln(β/T²) versus 1/T, the slope gives -E/R, where R is the gas constant (8.314 J mol⁻¹ K⁻¹). Also, the value of pre-exponential factor can be calculated from the intercept of the straight-line equation.

The OFW model is based on Doyle's approximation³² and the model is finally written by the equation.

\ln β = \ln (\frac{A_{o} E}{R g (α)}) - 5.331 - 1.052 \frac{E}{R T}

(11)

The slope of the plot ln β versus 1/T gives -E/R, which is employed to calculate kinetic parameters.

Continuous slope-based master plots

Master plots are useful in analyzing complex reaction mechanisms, such as biomass pyrolysis.^33–36 They provide a graphical representation of the reaction rate as a conversion function, allowing for identifying reaction pathways and mechanisms. It has been observed that master plots give more stable results in comparison with other iso-conversional models.³⁷ The master plots use differential and integral forms of the pre-defined mechanism in tandem with a set of kinetic equations recommended by the Committee of the International Confederation for Thermal Analysis and Calorimetry ICTAC,³⁸ as shown in Table 1. The present study used both the integral form of master plots g(α) and differential-integral coupling master plots Z(α). Mathematically, the basis for master plots is given for a solid-state reaction, working under the constrained environment of varying temperature, constant heating rate and constant reactor pressure. Determining functions g(α) and Z(α) were devised with novelty using the continuous slope method after defining equations (13) to (20). Figure 1 represents the algorithm for continuous slope-based master plot. The methodology used for the development of model is meticulously discussed in the following manner.

Figure 1.

Algorithm for continuous slope-based master plots.

Table 1.

Theoretical solid-state kinetic models along with their expressions.³⁸

Standard kinetic reaction model		g(α)	f(α)
F1	Reaction kinetic model	-ln(1 - α)	1 - α
F2		(1 - α)⁻¹ - 1	(1 - α)²
F3		[(1 - α)⁻² - 1]/2	(1 - α)³
D1	Gas diffusion model	α ²	1/2 α
D2		α + (1 - α)ln(1 - α)	[-ln(1 - α)]⁻¹
D3		[1 - (1 - α)^1/3]²	2/3(1 - α)^2/3/1 - (1 - α)^2/3
D4		1 - 2α/3 - (1 - α)^2/3	2/3(1 - α)^1/3/1 - (1 - α)^1/3
R1	Limiting surface reaction between both phases	α	1
R2		1 - (1 - α)^1/2	2(1 - α)^1/2
R3		1 - (1 - α)^1/3	3(1 - α)^2/3
A2	Nucleation growth	[-ln(1 - α)]^1/2	2(1 - α) [-ln(1 - α)]^1/2
A3	Nucleation growth	[-ln(1 - α)]^1/3	3(1 - α) [-ln(1 - α)]^2/3

The plot of g(α) is devised for both theoretical and experimental data points. Theoretically, the function $(d α / d T) / (d α / d T)_{0.5}$ is evaluated by employing the models of solid-state reaction as given in Table 1. The models are evaluated for the discrete values of conversion points in the range of 0.1–0.8. The experimental plots of $(d α / d T) / (d α / d T)_{0.5}$ were devised by a novel mathematical method. Firstly, the plot of α–T is obtained for all the values of α–T of TGA data at β = 50 °C min⁻¹. The empirical function (equation (12)) obtained was seen to be significantly continuous in the range, as shown in Figure 5, and thus differentiable within the given range. Mathematically, the empirical function was obtained after fitting the original curve with third degree polynomial with approximately 900 data points. The function developed is given as;

\frac{d α}{d T} = C_{0} + C_{1} T + C_{2} T^{2} + C_{3} T^{3}

(12)

Where the values of C₀, C₁, C₂, and C₃ are given as −0.12, 5.38E-4, −7.48E-7, and 3.35E-10, respectively. The value of coefficient of determination was found to be significantly higher as 0.98. The points showing infinite slope were rejected from the data, as shown in Figure 5. The continuous differential of the function α–T was obtained, and the value at α = 0.5 was noted. Secondly, the differential ratio for α = 0.1–0.8, with respect to α = 0.5, was evaluated at every point, and finally, it was plotted as experimental curve as shown in Figure 6(a) and (b). The comparison of models with respect to the experimental curve was done by using the Karl–Pearson correlation technique based on two-tailed tests of 0.05 level significance. The mathematical basis for the model proceeds as:

Rewriting equation (6), in terms of f(α), we get,

f (α) = \frac{d α}{d T} \times \frac{β}{A_{o}} e^{(E / R T)}

(13)

Equation (13), at the reference point of α = 0.5, gives

f (α)_{0.5} = {[\frac{d α}{d T}]}_{0.5} \times \frac{β}{A_{o}_{0.5}} \exp {(\frac{E}{R T})}_{0.5}

(14)

Dividing equations (13) by (14), we get

\frac{f (α)}{f (0.5)} = \frac{(d α / d T) \times \exp (E / R T) \times A_{o 0.5}}{{(d α / d T)}_{0.5} \times \exp {(E / R T)}_{0.5} \times A_{o}}

(15)

where the term

(d α / d T)_{0.5}

denotes the slope of α–T plot at α = 0.5, f(0.5), T_0.5, and E_0.5 denote the function value, temperature value, and apparent activation energy value at α = 0.5, respectively.

By assuming the value of the ratio of the pre-exponential factor to unity, equation (15) can be further reduced to equation (16).

\frac{f (α)}{f (0.5)} = \frac{(d α / d T) \times \exp (E / R T)}{{(d α / d T)}_{0.5} \times \exp {(E / R T)}_{0.5}}

(16)

The LHS of equation (16) plots theoretical master curves based on f(α), employing various solid-state kinetic models as given in Table 1. These theoretical plots between f(α) and α are compared with the experimental master curve, plotted with the help of RHS of equation (16). The standard correlation technique is used to compare the goodness of fit.

Similarly, the master plots based on both differential and integral solid-state models f(α) and g(α) are combined by unified function Z(α).

Z (α) = f (α) * g (α)

(17)

Where, g(α) is given by:

g (α) = \int_{0}^{\infty} \frac{d α}{f (α)}

(18)

From equations (6) and (18)

g (α) = \int_{T \circ}^{T} \frac{A_{o}}{β} \exp (- \frac{E}{R T}) d T

(19)

Multiplying equations (13) and (19), we get Z(α)

Z (α) = \frac{d α}{d T} * \exp (\frac{E}{R T}) * \int_{T \circ}^{T} (\frac{- E}{R T}) d T

(20)

The LHS of equation (20) is used to plot theoretical master plots based on Z(α), employing various solid-state reaction kinetic models as given in Table 1. Again, these theoretical plots are compared with the experimental master plot curve, which is plotted by employing the RHS of equation (20). The value of apparent activation energy (E) and corresponding temperature (T) are used at heating rate of 50 °C min⁻¹. The value of

d α / d T

is calculated from the slope of the α–T curve at different values of α. Finally, the experimental curve is compared with other standard curves by estimating the goodness of fit parameter.

Statistical analysis

The robustness of any model is determined by its statistical metrics. In this study the statistical analysis used to validate and test the coherence of various pre-defined models with experimentally determined mechanisms was done using a built-in function (Karl–Pearson correlation) in Origin 2021 software. The curve-fitting data analysis was conducted using the Karl–Pearson package provided by the software. Adjusted R² were chosen for determining the goodness of fit. The higher values of R² (>0.75) were considered statistically fit for the comparison. The p-value of (p < 0.005) was considered statistically significant.

Result and discussions

Physicochemical characterization of rice husk

The value of key parameters during the proximate analysis of RH showed a moisture content (11.57 wt-%), volatile matter (54.51 wt-%), fixed carbon (10.24 wt-%), and ash (23.68 wt-%) on dry basis. The data generated is in line with the presented average data of rice husk as discussed in Table 2. The lower moisture content reveals higher heating value in addition to increase in conversion efficiency. It should be noted that despite its relatively high ash content, rice husk still fares better than other waste and solid fuels commonly used for energy production, which often exceed ash content levels of 15%. However, the high ash content of rice husk can lead to decreased heating value and furnace clogging during thermal conversion under high temperatures.³⁹ Conversely, the medium levels of volatile matter and fixed carbon content in rice husk are consistent with those found in terrestrial biomass. The possible reason underlying the fact of medium values of volatile matter and fixed carbon is due to high silica content and lower bulk density of rice husk. On the other hand, the sample contained 40.23 wt-%, 4.516 wt-%, and 46.13 wt-% of carbon, hydrogen, and oxygen, respectively. The gross calorific value (HHV) of RH sample was found to be 18.42 MJ kg⁻¹ as theoretically calculated by equation (1), while the experimental value was around 17.31 MJ kg⁻¹ and thus the values are in close association with the study carried.⁴⁰ The medium value of HHV can be attributed to higher ash formation along with the higher silica content. The production of gases like nitrogen and sulfur was seen significantly lower than 1.6 wt-% which hints towards the favorable syngas production along with pyrolysis oil formation.⁴¹ The energy content of biomass is less when used directly than the converted one. This is because C–C bonds have higher energy content than C–O and C–H bonds and the biomass is rich in oxygen than carbon.⁴² This difference in composition results in low carbon, hydrogen, and oxygen content and relatively higher nitrogen and sulfur content compared to other lignocellulosic biomasses.⁴³

Table 2.

Consolidated characteristics of RH samples.

Proximate analysis
Property	% (dry basis)	%age range (dry basis) from literature	Source
Moisture content	11.57	7.14–16.54	^31,40,44
Ash content	23.68	5.87–14.56
Volatile matter	54.51	65.20–70.08
Fixed carbon	10.24	8.02–12.05

Ultimate analysis
Element	% (dry basis)	%age range (dry basis) from literature	Source
Carbon	40.23	38.20–44.00	^31,40,44
Hydrogen	4.516	4.98–5.205
Nitrogen	1.513	0.63–1.40
Sulfur	0.180	0.19–1.01
Oxygen	46.13	47.68–55.53
HHV (MJ kg⁻¹)	Experimental	17.31
HHV (MJ kg⁻¹)	Theoretical	18.42

Pyrolysis behavior for TGA analysis

The behavior of pyrolysis was studied by the applications of TGA and DTG. The TGA experiment imitates the pyrolytic conditions at pilot scale where the weight loss of sample is observed as a function of temperature or time.⁴⁵ The TG and DTG curves for rice husk at heating rates of 20, 50, and 100 °C min⁻¹ are presented in Figure 2. The TG and DTG curves revealed that the thermal behavior of rice husk follows three different stages of weight loss including the drying process between room temperature and 225 °C. During thermal decomposition in an inert atmosphere, the chemical bonds of the proteins, lipids, carbohydrates, and other compounds in RH break down into light hydrocarbons, including bio-syngas and bio-oil products.⁴⁶ Table 3 summarizes the significant weight losses during pyrolytic stages. Stage-I could be attributed to the process of drying, in which most of the moisture upto the equilibrium is released. The weight loss percentage in stage-I, when the temperature rises from 20 to 225 °C, is 10.9 wt-%. The moisture or volatile substances contained in the sample evaporates or volatilizes during this initial phase. The change in slope from being too steep to flat can be attributed to intrinsic drying characteristics of a rice husk. The removal of unbounded moisture occurs at higher rate and thus the slope appears steep whereas for bounded moisture the drying rate is relatively slower. The stage-I is for devolatilization of cellulosic and hemicellulosic components of biomass.¹¹ This stage is prominently identified as the complex stage of pyrolysis due to release of many volatile substances as the function of temperature.⁴⁷

Figure 2.

TGA and DTG characteristics of rice husk at three different heating rates.

Table 3.

Weight loss in different stages of pyrolysis of RH as a function of temperature.

Heating rates (°C min⁻¹)	Stage-I			Stage-II			Stage-III
Heating rates (°C min⁻¹)	T_i (°C)	T_f (°C)	Weight loss (%)	T_i (°C)	T_f (°C)	Weight loss (%)	T_i (°C)	T_f (°C)	Weight loss (%)
20	0	225	10.9	225	410	50.45	410	576	24.65
50	0	226	8.3	226	395	44.3	395	729	30.2
100	0	226	8.1	226	485	50.36	485	916	25.04

In stage-II, which lasts from 225 to 480 °C, the proportion of weight lost jumps to a startling 50.45 wt-%. The sample undergoes a more sudden and complicated decomposition process involving disintegration of organic molecules, polymers, or other heat-sensitive components like CO₂, H₂, N₂, and CO as suggested by the sample's large weight loss. The weight loss shows steep gradient and the predominant peaks were observed around 250 and 480 °C. Stage-III is the last stage comprising of lignin decomposition with temperatures ranging from 480 to 950 °C with a weight loss of 24.65 wt-%. This phase could mark the culmination of the breakdown of heat-sensitive chemicals, which would result in the creation of stable residual char or ash.⁴⁸ The appendage of long tails in the plot shows the formation of ash by degradation of inorganics present in the RH.⁴⁹ Thus, the detailed analysis of three different stages concludes stage-II as the active zone of pyrolysis with temperature range of 225–480 °C. The zone stands most important from the industrial point of view as the major decomposition take place in this zone.

The curves of TGA and DTG clearly show a shift towards higher peak temperature. The peak temperature of DTG curves for β = 20, 50, 100 °C min⁻¹ are 351, 355, and 360 °C, respectively. It was also seen that the peak temperatures were elevated with corresponding increase in the heating rates. The delay in the elevated peaks observed were as a result of detrimental effects of shorter residence time complementing the reduction in effective heat transfer.⁵⁰ Table 3 illustrates the effect of heating rates on overall degradation along with percentage weight loss. The stage-I and stage-II can be seen following the rule of thumb of higher weight loss at lower heating rates. The lower heating rates give ample amount of time for degradation process at particular temperature to take place. Conversely, the stage-III shows lower weight loss at lower heating rate. The anomaly may be attributed to measurement errors arising due to further reduction in weight of remaining residue. The material's thermal stability and reaction kinetics can be learned a lot from the DTG peak temperatures. The derivative of the weight loss with respect to time determines the maximum rate of weight loss during thermal decomposition, which corresponds to these peak temperatures. The reaction kinetics is dependent on the heating rate because the peak temperature shifts to a higher temperature as the rate of heating increases.⁵¹ The Arrhenius equation, which describes the temperature dependence of the reaction rate provides sufficient explanation for the dependence of reaction kinetics on the heating rate. The heating rate in DTG analysis can be thought of as a controlled rise in temperature that has a direct impact on the reaction rate. The rate of energy transfer to the sample increases with the heating rate, resulting in a faster reaction rate.⁵² According to our findings, the shift in peak temperature to higher values authenticates our study as well.

Kinetic modelling

The kinetic parameters, namely apparent average activation energy (E) and pre-exponential factor (A_o), were determined by employing KAS and OFW iso-conversional models along with curve-fitting technique. The conversion points (α) were used discretely in the range of 0.1 ≤ α ≤ 0.8, as shown in Table 4. Figure 3 shows the straight-line fit of both KAS and OFW models, respectively at progressive conversions. With the increase in conversion, the slope of the fitted lines changed in both KAS and OFW models which shows the non-linear dependencies of the process on conversion. The apparent activation energy values calculated, range from 161.973 to 259.738 kJ mol⁻¹ for the KAS model and from 150.861 to 259.738 kJ mol⁻¹ for OFW model. The average values of apparent activation energy for both the KAS model and OFW model were found to be 187.29 and 199.85 kJ mol⁻¹, respectively. The corresponding average R² values were found to be 0.988 and 0.882, respectively. Few studies reported a similar result that the apparent activation energy of RH ranges from 190.8 to 201 kJ mol⁻¹.^48,53 The values indicate the energy required for the decomposition process to occur. The higher the activation energy, the more energy is needed for the reaction to proceed. The average value of apparent activation energy was somehow comparable, with an appreciable increase for the OFW model, as depicted in Figure 4 as well. The possible reason underlying this seems to be the different approximation of temperature integral in determining the mechanism of pyrolysis. The fluctuation in apparent activation energy during rice husk pyrolysis demonstrates a complex relationship between biomass composition and reaction kinetics. The apparent activation energy drops first up to conversion point α = 0.6, indicating the quick breakdown of weaker bonds in cellulose and hemicellulose. Another reason could be attributed to the relative lower activation energy value of cellulose along with its higher initial weight loss. The cumulative effect of the change in the apparent activation energies of cellulose and hemi cellulose thus shows decreasing trend up to α = 0.6. Once above 0.6, the apparent activation energy increases dramatically due to the accumulation of refractory components in the residual biomass, necessitating more energy to disrupt their robust connections.⁴⁶ This non-linear variation of apparent activation energy with respect to conversion during pyrolysis of rice husk also suggests that the reaction kinetics change as the conversion proceeds. This could be due to various factors, such as the formation of different reaction intermediates, changes in the reaction mechanism, or the influence of temperature and other environmental conditions.^38,54 The different composition (RH consists of 33% cellulose, 26% hemicellulose, and 7% lignin) in the RH feedstock results in the variation of apparent activation energy and studies have revealed that hemicellulose has a higher activation energy than cellulose because cellulose has a well-organized crystal structure, which leads to a considerably easier degradation pathway than hemicellulose.⁴⁶ Additionally, the main reaction for hemicellulose and lignin during the pyrolysis process at high conversion rates is condensation, which progresses at a relatively sluggish rate, leading to a considerable increase in apparent activation energy.⁵⁵

Figure 3.

Linear fit to determine kinetic parameters of RH using (a) KAS model and (b) OFW model through progressive conversion points.

Figure 4.

Variation of apparent activation energy with progressive conversion.

Reaction mechanism and its interpretation by master plots

Figure 6 shows the master plot curves of g(α) and Z(α). The master plot method demonstrates the complex mechanism of the pyrolytic degradation process of rice husk, which can be clearly seen due to multiple higher correlation coefficients arising during comparison of experimental curve with solid-state reaction equations. As observed from Table 5, the experimental curve had the highest correlation coefficient of 0.94 and 0.75 with respect to A2 and A3 models (Avrami–Erofeev), respectively, for g(α). Thus, it can be inferred that the nucleation models are dominant for a larger fraction of reaction path. The reason for the Avrami–Erofeev effect could be attributed to the intrinsic cracks and defects of the sample which act as nucleation sites.⁵⁶ For Z(α), the higher values of Pearson's coefficient was found for F3 and D1 models, with values of 0.8 and 0.81, respectively. The association hints that the process was limited by first-order multi-diffusional processes coupled with solid phase surface reaction. Moreover, the reason for the mechanism could be attributed to the formation of char with outpouring of condensable and non-condensable gases at its surface. It was also observed that the mechanism of degradation of RH was highly correlated with the progression of conversion. The diffusion model D1 was seen significantly associated with the experimentally determined model in the span of α = 0.1 to α = 0.5. Above conversion of 0.5 all the three models except D1 were closely associated with the experimental model. Also, the master plots from experimental data infers the generalized model of f(α) as $f (α) = 12.74 α^{3} - 24 α^{2} + 12.6 α - 0.92$ . The R² value was found to be significantly high as 0.996. Thus, it was concluded that for differential models, the mechanism fits with Avrami–Erofeev nucleation models, while as, for combined integral and differential models the mechanism fits with multi-diffusional processes coupled with solid phase surface reaction models. The similar trends have been reported for other biomasses as well,^47,57,58 thus it may be settled that the mechanism of pyrolytic degradation of rice husk follows Avrami–Erofeev nucleation and multi-diffusion model due to a multiple parallel reaction kinetics.⁵⁹

Figure 5.

Continuous slope-based empirical function (dα/dT) with respect to temperature (T).

Figure 6.

Experimental and theroetical master plots of RH: for (a) g(α) and (b) Z(α).

Table 4.

Apparent activation energy and pre-exponential factor of RH at different conversion points using KAS and OFW models.

Conversion	KAS model			OFW model
	E (kJ mol⁻¹)	A_o (s⁻¹)	R ²	E (kJ mol⁻¹)	A_o (s⁻¹)	R ²
0.1	245.192	9.0E + 14	0.964	259.738	6.3E + 17	0.663
0.2	236.056	4.2E + 13	0.969	245.352	5.9E + 18	0.748
0.3	205.247	6.9E + 14	0.990	196.36	5.2E + 18	0.875
0.4	190.989	2.8E + 13	0.998	185.398	4.2E + 18	0.909
0.5	145.562	5.2E + 12	0.999	171.774	2.0E + 17	0.948
0.6	142.250	1.5E + 15	0.998	170.544	7.5E + 18	0.999
0.7	161.070	1.2E + 20	0.995	182.861	4.5E + 19	0.923
0.8	171.973	8.4E + 18	0.992	186.824	5.5E + 23	0.994
Average	187.290	1.6E + 19		199.850	6.88E + 22

Table 5.

Statistical analysis of different standard models in comparison with experimental master plots.

Model		F1	F2	F3	D1	D2	D3	D4	R2	R3	A2	A3
g(α)	Pearson correlation	0.2	0.06	−0.07	−0.2	−0.3	−0.28	−0.27	0.3	0.28	0.94	0.75
Z(α)	Pearson correlation	0.27	0.64	0.8	0.8	−0.05	0.37	0.52	0.26	0.45	0.52	0.67

Conclusions

Following an extensive physiochemical characterization of locally available RH, the pyrolytic degradation kinetics was studied through TG/DTG at three heating rates of 20, 50, and 100 °C min⁻¹. The peak temperature values of DTG showed an active pyrolysis zone between 250 and 550 °C, with the degradation percentage of about 60% by weight. The probable reaction mechanism was studied by novel model-free method of continuous slope-based master plots in addition to the KAS and OFW models. The average apparent activation energy was found to be 182.79 and 199.85 kJ mol⁻¹ for KAS and OFW model, respectively, following multiple reaction mechanisms during pyrolysis.

Footnotes

Author contributions

Ishfaq Najar did writing-original draft, formal analysis, data generation, and validation.

Tanveer Rasool did conceptualization, draft checking, data curation, and supervision.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Ishfaq Najar

Tanveer Rasool

Ishfaq Najar is PhD scholar at the Department of Chemical Engineering, National Institute of Technology Srinagar-India. He holds a Masters degree in Chemical Engineering from National Institute of Technology Srinagar-India. His research areas cover, among others, Biomass Conversion Technology, Biofuels, Modeling and Simulation, and Machine learning.

Tanveer Rasool is an Assistant Professor at the Department of Chemical Engineering, at National Institute of Technology Srinagar-India. He holds a PhD in Chemical Engineering from National Institute of Technology Srinagar-India. His research areas cover, among others, Biomass Conversion Technology, Industrial Pollution Abatement, Modeling and Simulation, Material Technology, bio-energy and Environment Technology development.

References

Bouckaert

Fernandez Pales

McGlade

, et al. Net zero by 2050—analysis—IEA. International Energy Agency (IEA) 2021: 1–224.

Pandey

Sirohi

, et al. Precise prediction of biochar yield and proximate analysis by modern machine learning and SHapley Additive exPlanations. Energy Fuels 2023; 37: 17310–17327.

PwC. Fuelling India’s future with bioenergy; 2023: 1–15.

Amoloye

Abdulkareem

Adeniyi

. Thermo-kinetics, thermodynamics, and ANN modeling of the pyrolytic behaviours of Corn Cob, Husk, Leaf, and Stalk using thermogravimetric analysis. Chem Prod Process Model 2023; 18: 859–876.

India: production volume of rice 2023 | Statista, https://www.statista.com/statistics/1140236/india-production-volume-of-rice/#statisticContainer (accessed 16 January 2024).

Herrera

Morales

Tarazona

, et al. Use of biochar from rice husk pyrolysis: part A: recovery as an adsorbent in the removal of emerging compounds. ACS Omega 2022; 7: 7625–7637.

Chindaprasirt

Kanchanda

Sathonsaowaphak

, et al. Sulfate resistance of blended cements containing fly ash and rice husk ash. Constr Build Mater 2007; 21: 1356–1361.

Rather

. Production and productivity trends of paddy cultivation in Jammu & Kashmir. Paripex – Indian J Res 2014; 3: 42–44.

Liu

van Groenigen

, et al. Rice paddy soils are a quantitatively important carbon store according to a global synthesis. Commun Earth Environ 2021; 2: 1–9.

10.

Hoang

Ong

Fattah

IMR

, et al. Progress on the lignocellulosic biomass pyrolysis for biofuel production toward environmental sustainability. Fuel Process Technol 2021; 223: 106997.

11.

Rajamohan

Chidambaresh

Sundarrajan

, et al. Investigation of thermodynamic and kinetic parameters of Albizia lebbeck seed pods using thermogravimetric analysis. Bioresour Technol 2023; 384: 129333.

12.

Tagade

Kandpal

Sawarkar

. Insights into pyrolysis of pearl millet (Pennisetum glaucum) straw through thermogravimetric analysis: physico-chemical characterization, kinetics, and reaction mechanism. Bioresour Technol 2024; 391: 129930.

13.

Marques J de

Alves

JLF

de Oliveira

KFS

, et al. Catalytic upgrading of pyrolysis vapors from Scenedesmus sp. microalgae towards renewable hydrocarbons using a low-cost zeolite synthesized from rice husk ash and diatomite residue. Bioenergy Res. Epub ahead of print 19 February 2024. DOI: https://doi.org/10.1007/S12155-024-10732-9

14.

Wakatuntu

Olupot

Jjagwe

, et al. Optimization of pyrolysis conditions for production of rice husk-based bio-oil as an energy carrier. Results Eng 2023; 17: 100947.

15.

Olupot

Wakatuntu

Turyasingura

, et al. Optimization of heavy metal removal by activated carbon obtained as a co-product from fast pyrolysis of rice husks. Results Mater 2024; 21. DOI: https://doi.org/10.1016/j.rinma.2024.100545

16.

Téllez

Silva

Simister

, et al. Fast pyrolysis of rice husk under vacuum conditions to produce levoglucosan. J Anal Appl Pyrolysis 2021; 156: 105105.

17.

Pawar

Panwar

Jain

, et al. Thermal degradation of coconut husk waste biomass under non-isothermal condition. Biomass Convers Biorefin 2023; 13: 7613–7622.

18.

Singh

Patil

Sawarkar

. Pyrolysis of garlic husk biomass: physico-chemical characterization, thermodynamic and kinetic analyses. Bioresour Technol Rep 2020; 12: 100558.

19.

Cai

Dai

Liu

. Catalytic fast pyrolysis of rice husk for bio-oil production. Energy 2018; 154: 477–487.

20.

Zhang

Dong

Zhang

, et al. High quality syngas production from microwave pyrolysis of rice husk with char-supported metallic catalysts. Bioresour Technol 2015; 191: 17–23.

21.

Sahoo

Remya

. Influence of operating parameters on the microwave pyrolysis of rice husk: biochar yield, energy yield, and property of biochar. Biomass Convers Biorefin 2022; 12: 3447–3456.

22.

Potnuri

Suriapparao

Rao

, et al. Understanding the role of modeling and simulation in pyrolysis of biomass and waste plastics: a review. Bioresour Technol Rep 2022; 20: 101221.

23.

Rahbari

Ebrahiminasab

Bidabadi

. Analytical model for predicting the heat loss effect on the pyrolysis of biomass particles. Chin J Chem Eng 2013; 21: 1114–1120.

24.

Ling

CCY

SFY

. Synergistic interactions between sewage sludge, polypropylene, and high-density polyethylene during co-pyrolysis: an investigation based on iso-conversional model-free methods and master plot analysis. J Hazard Mater 2023; 455: 131600.

25.

Nie

Wang

Yellezuome

, et al. Exploring kinetic and thermodynamic mechanisms of switchgrass pyrolysis using iterative linear integral isoconversional method and master plots approach. Fuel 2023; 338: 127266.

26.

Agnihotri

Mondal

. Thermal analysis, kinetic behavior, reaction modeling, and comprehensive pyrolysis index of soybean stalk pyrolysis. Biomass Convers Biorefin 2023; 1: 1–16.

27.

Ram Nhuchhen

Abdul Salam

. Estimation of higher heating value of biomass from proximate analysis: a new approach. Fuel 2012; 99: 55–63.

28.

ASTM D-00. Standard test method for gross calorific value of coal and coke by the adiabatic bomb calorimeter (withdrawn 2000). ASTM International. https://www.astm.org/ (2000, accessed 13 June 2023).

29.

Arenas

Navarro

Martínez

. Pyrolysis kinetics of biomass wastes using isoconversional methods and the distributed activation energy model. Bioresour Technol. Epub ahead of print 2019: 288. DOI: https://doi.org/10.1016/j.biortech.2019.121485

30.

Rasool

Najar

Srivastava

, et al. Pyrolysis of almond (Prunus amygdalus) shells: kinetic analysis, modelling, energy assessment and technical feasibility studies. Bioresour Technol 2021; 337: 125466.

31.

Lim

ACR

Chin

BLF

Jawad

, et al. Kinetic analysis of rice husk pyrolysis using Kissinger–Akahira–Sunose (KAS) method. Procedia Eng 2016; 148: 1247–1251.

32.

Doyle

. Kinetic analysis of thermogravimetric data. J Appl Polym Sci 1961; 5: 285–292.

33.

Luo

Guo

Zhang

, et al. Insight into pyrolysis kinetics of lignocellulosic biomass: isoconversional kinetic analysis by the modified Friedman method. Energy Fuels 2020; 34: 4874–4881.

34.

Wang

, et al. Kinetic study of the effect of in-situ mineral solids on pyrolysis process of oil sludge. Chem Eng J 2019; 374: 338–346.

35.

Sarbishei

Tafaghodi

. A comprehensive kinetic investigation on sulfur deportment during nickel laterite calcination using model-free, model-fitting, and master plot approaches. Miner Eng 2023; 192: 108001.

36.

Ren

Zhou

, et al. The influence of sludge ash and its active species on the combustion characteristics and kinetics of biomass: application of model-fitting method combined with master-plots method. Biomass Convers Biorefin. Epub ahead of print 2023: 192. DOI: https://doi.org/10.1007/s13399-023-04308-4

37.

Pérez-Maqueda

Criado

Gotor

, et al. Advantages of combined kinetic analysis of experimental data obtained under any heating profile. J Phys Chem A 2002; 106: 2862–2868.

38.

Vyazovkin

Burnham

Criado

, et al. ICTAC Kinetics Committee recommendations for performing kinetic computations on thermal analysis data. Thermochim Acta 2011; 520: 1–19.

39.

Singh

Gokul

, et al. Thermal degradation and pyrolysis kinetics of two Indian rice husk varieties using thermogravimetric analysis. Energy Sources A: Recovery Util Environ Effects 2020; 520(1-2): 1–12.

40.

Bhattacharjee

Biswas

. Catalytic pyrolysis of rice husk with SnCl₂, Al₂O₃.4SiO₂.H₂O, and MoS2 for improving the chemical composition of pyrolysis oil and gas. J Indian Chem Soc. Epub ahead of print 2022: 99. DOI: https://doi.org/10.1016/j.jics.2022.100728

41.

Friedl

Padouvas

Rotter

, et al. Prediction of heating values of biomass fuel from elemental composition. Anal Chim Acta 2005; 544: 191–198.

42.

Tahir

Zhao

Ren

, et al. Thermo-kinetics and gaseous product analysis of banana peel pyrolysis for its bioenergy potential. Biomass Bioenergy 2019; 122: 193–201.

43.

Liu

Wang

Karim

, et al. Catalytic fast pyrolysis of lignocellulosic biomass. Chem Soc Rev 2014; 43: 7594–7623.

44.

Yousefian

Babatabar

Eshaghi

, et al. Pyrolysis of rice husk, coconut shell, and Cladophora glomerata algae and application of the produced biochars as support for cobalt catalyst in Fischer–Tropsch synthesis. Fuel Process Technol 2023; 247: 107818.

45.

Escalante

Chen

Tabatabaei

, et al. Pyrolysis of lignocellulosic, algal, plastic, and other biomass wastes for biofuel production and circular bioeconomy: a review of thermogravimetric analysis (TGA) approach. Renewable Sustainable Energy Rev 2022; 169: 112914.

46.

Chen

Cen

Zhuang

, et al. Insight into biomass pyrolysis mechanism based on cellulose, hemicellulose, and lignin: evolution of volatiles and kinetics, elucidation of reaction pathways, and characterization of gas, biochar and bio-oil. Combust Flame 2022; 242: 112142.

47.

Rasool

Srivastava

Khan

MNS

. Utilisation of a waste biomass, walnut shells, to produce bio-products via pyrolysis: investigation using ISO-conversional and neural network methods. Biomass Convers Biorefin 2018; 8: 647–657.

48.

Zhang

Wang

, et al. Pyrolysis characteristics, kinetics and evolved volatiles determination of rice-husk-based distiller’s grains. Biomass Bioenergy 2020; 135: 105525.

49.

Saadatkhah

Carillo Garcia

Ackermann

, et al. Experimental methods in chemical engineering: thermogravimetric analysis—TGA. Can J Chem Eng 2020; 98: 34–43.

50.

Petrovič

Stergar

Škodič

, et al. Thermo-kinetic analysis of pyrolysis of thermally pre-treated sewage sludge from the food industry. Therm Sci Eng Progress 2023; 42: 101863.

51.

Ahmad

Mehmood

, et al. Thermogravimetric analyses revealed the bioenergy potential of Eulaliopsis binata. J Therm Anal Calorim 2017; 130: 1237–1247.

52.

Ceylan

Topçu

. Pyrolysis kinetics of hazelnut husk using thermogravimetric analysis. Bioresour Technol 2014; 156: 182–188.

53.

Loy

ACM

Gan

DKW

Yusup

, et al. Thermogravimetric kinetic modelling of in-situ catalytic pyrolytic conversion of rice husk to bioenergy using rice hull ash catalyst. Bioresour Technol 2018; 261: 213–222.

54.

Kaur

Gera

Jha

, et al. Pyrolysis kinetics and thermodynamic parameters of castor (Ricinus communis) residue using thermogravimetric analysis. Bioresour Technol 2018; 250: 422–428.

55.

Chen

Wei

Zhao

, et al. Study on co-pyrolysis and products of Chlorella vulgaris and rice straw catalyzed by activated carbon/HZSM-5 additives. Bioresour Technol 2022: 360. DOI: https://doi.org/10.1016/j.biortech.2022.127594

56.

Vasudev

Lin

. Pyrolysis of algal biomass: determination of the kinetic triplet and thermodynamic analysis. Bioresour Technol 2020; 317: 124007.

57.

Luo

Zhang

, et al. Insight into master plots method for kinetic analysis of lignocellulosic biomass pyrolysis. Energy 2021; 233: 121194.

58.

Patidar

Singathia

Vashishtha

, et al. Investigation of kinetic and thermodynamic parameters approaches to non-isothermal pyrolysis of mustard stalk using model-free and master plots methods. Mater Sci Energy Technol 2022; 5: 6–14.

59.

Liu

. Pyrolysis characteristic and kinetic analysis of sewage sludge using model-free and master plots methods. Process Saf Environ Prot 2021; 149: 48–55.