The past decade has seen increasing interest in production of fuels and chemicals from biomass. Based on the types of feedstock used, biofuels are classified as either first or second-generation.¹  First-generation biofuels include ethanol produced from sugars and starch crops such as maize and sugarcane and biodiesel from seed oils. In contrast, second-generation biofuels are produced from cellulosic and lipid-rich plant materials that are not food crops. These include agricultural and forestry residues, dedicated energy crops like hybrid poplar and switchgrass, algae and municipal solid waste. Although the commercial production of first-generation biofuels has grown tremendously in the last decade, they have been challenged for their limited greenhouse gas reductions compared to petroleum-based fuels and concerns that their production diverts these crops from food production, the so-called food-vs.-fuel debate.¹  Second-generation biofuels offer the prospect of overcoming both of these challenges compared to first-generation biofuels.

Second-generation biofuels can be produced by thermochemical or biochemical processes.²  Thermochemical processing utilizes heat and catalysts while biochemical processing employs enzymes and microorganisms to convert biomass into fuels and chemicals. Hybrid processing, which combines aspects of thermochemical and biochemical processing, is of growing interest.^3,4  Figure 1.1 summarizes the conversion of cellulosic and lipid-rich feedstocks by biochemical and thermochemical processes into diverse products.

Thermochemical processes, operating at significantly higher temperatures than biochemical processing, are usually very fast, measured in seconds or minutes compared to hours or days for biochemical processes. On the other hand, thermochemical processes can be less selective than biochemical processes, which can unfavorably affect yields of desired products. However, this lack of selectivity often means that more kinds of feedstock molecules are converted, resulting in higher overall yields of drop-in fuels from lignocellulosic feedstocks, for example.

Thermochemical processes can be classified into gasification, pyrolysis, and solvent liquefaction.⁵  Gasification converts solid feedstocks into flammable gases known as producer gas or syngas. Pyrolysis converts solid feedstocks into mostly liquid products. Solvent liquefaction resembles pyrolysis in some respects, producing mostly liquid products, but occurs in the presence of a solvent. Of these three thermochemical technologies, pyrolysis has received the most attention in the last few years for its potential to convert lignocellulosic biomass into a liquid intermediate that can be upgraded to drop-in (hydrocarbon) fuels using technologies familiar to the petroleum industry. It also has prospects for distributed processing of biomass, which can simplify the logistics of providing feedstock to a processing plant.^6,7 

Pyrolysis is the thermal decomposition of organic substances in the absence of oxygen to form liquids, solids, and non-condensable gases. The rate of pyrolysis profoundly affects product distributions. Slow pyrolysis, developed centuries ago to produce charcoal for heating purposes, occurs over periods measured in hours or even days. In contrast, fast pyrolysis both rapidly heats the feedstock and quenches the products, usually in the order of seconds, with the goal of producing an energy-rich liquid, known as bio-oil, from the vapors as the primary product. Although originally produced for use as heating oil or electric power generation, bio-oil has been increasingly regarded as an intermediate for the production of drop-in biofuels, biobased chemicals, and hydrogen fuel. To maximize bio-oil production (up to 75 wt% of biomass), several conditions must be met during pyrolysis:^8,9 

• the biomass must be rapidly heated, in the order of a few seconds;

• the products of pyrolysis must be rapidly removed from the reaction zone and cooled, in the order of a few seconds;

• optimum reaction temperature is thought to be between 400–500 °C.

The solid product of fast pyrolysis, known as biochar, consists mostly of carbon but also contains ash originating from biomass. Biochar, which represents 12–15 wt% of the products of fast pyrolysis, can be used as boiler fuel but more intriguing applications include soil amendment, carbon sequestration agent, and activated carbon.¹⁰  Non-condensable gases from fast pyrolysis, yielding 13–25 wt%, are a flammable mixture of carbon monoxide, hydrogen, carbon dioxide, and light hydrocarbons suitable for generating process heat.⁹ 

Many kinds of lignocellulosic biomass, ranging from agricultural residues, forestry waste, and energy crops, have been tested for suitability as fast pyrolysis feedstock. The three major components of lignocellulosic biomass, illustrated in Figure 1.2, are cellulose, hemicellulose, and lignin. Cellulose, the most abundant polymer on the planet, constitutes 30–50% of lignocellulosic biomass. It is a structural polysaccharide consisting of pyranose rings linked by glycosidic bonds. Hemicellulose is a heteropolysaccharide of random, amorphous structure cross-linked to cellulose and lignin. Lignin is a highly branched phenol-based polymer bound to cellulose and lignin to form a lignocellulosic matrix.²  Each of these components produce distinctive products under fast pyrolysis.

Fast pyrolysis has also been used to thermally deconstruct other kinds of biomass feedstocks such as algae and a variety of mixed wastes including municipal solid waste, sewage sludge, manure, food processing waste, and organic by-products from manufacturing.^9,11,12  These feedstocks often contain relatively large fractions of starch, lipids and proteins compared to lignocellulosic biomass, as well as significant ash. The compositional complexities of alternative feedstocks adds to the difficulties of pyrolyzing them.^13,14 

Drying biomass feedstocks to less than 10 wt% moisture and comminution to less than 3 mm particle diameter are important steps to successful fast pyrolysis. Despite this preprocessing, the low bulk density, irregular particle shapes and cohesive/adhesive behavior of many kinds of biomass can lead to bridging, blockage, and other feeding difficulties.¹⁵ 

Although fast pyrolysis was first investigated as early as 1875,¹⁶  significant progress in developing it for bio-oil production only dates from the 1980s. A variety of reactors were investigated with the goal of heating biomass to temperatures exceeding 400 °C in a few seconds. Suitable reactors include bubbling fluidized beds, circulating fluidized beds, rotating cone reactors, auger reactors, entrained flow reactors and ablative reactors.^8,17  Key features of these different classes of fast pyrolysis reactors are summarized in Table 1.1.

Among the various kinds of reactors, fluidized beds have received the most attention for fast pyrolysis due to excellent heat and mass transfer characteristics, simplicity of operation, and relative ease of scale-up. Bio-oil yields of 60–75% from fast pyrolysis of lignocellulosic biomass have been reported.⁸  Canadian-based Dynamotive commercialized the bubbling fluidized bed fast pyrolysis technology with a capacity of 200 tons of biomass per day (ton day⁻¹).¹⁸  Another Canadian company, Ensyn Technologies Inc. has developed and commercialized its Rapid Thermal Processing (RTP) process since the early 1980s, which is based on circulating fluidized bed technology. Several RTP™ biomass pyrolysis plants are in commercial operation, with capacity ranging from 1 to 200 ton day⁻¹.¹⁸ 

Rotating cone reactors were originally developed at Twente University, with the aim of eliminating carrier gas requirements while maintaining a high throughput and rapid heat transfer similarly to fluidized bed reactors. Instead of using inert gas, biomass is mechanically mixed with a solid heat transfer medium in a rotating cone. The rotating cone reactor concept has been commercialized by Biomass Technology Group (BTG) in the Netherlands, building several reactors with capacities of up to 120 ton day⁻¹.^8,18  The auger reactor, which also features mechanical mixing of biomass with a solid heat transfer medium, has been developed by ABRI-Tech of Canada and the Karlsruhe Institute of Technology (KIT). A pilot scale twin-screw auger reactor with 12 ton day⁻¹ capacity is now operational at KIT.¹⁹  Other types of reactor, including entrained flow, vacuum moving bed, ablative reactor, and microwave pyrolysis reactors are still in relatively early stage of technology development.

Quenching of the condensable vapors and aerosols from fast pyrolysis produces bio-oil. The rapid removal and cooling of pyrolysis products to minimize secondary reactions that can crack vapors are critical to achieving high liquid yields. Traditionally this is accomplished by spraying cold liquid hydrocarbon or recycling bio-oil into the pyrolysis vapor stream. The bio-oil collected in this fashion is an emulsion of lignin-derived phenolic compounds in an aqueous phase containing mostly carbohydrate-derived compounds.¹⁸  Unfortunately, the high reactivity of the bio-oil makes it difficult to distill into separate components, limiting opportunities for upgrading it to diverse value-added products. More recently, efforts have been made to develop fractionating bio-oil collection systems based on boiling points of bio-oil constituents. One such system consists of pairs of temperature-controlled condensers and electrostatic precipitators operated in series to remove vapors and the aerosols as heavy ends (anhydrosugars and phenolic oligomers), a middle fraction (furans and phenolic monomers), and light ends (water and light oxygenates).²⁰ 

Although the dark-brown liquid collected from fast pyrolysis is called “bio-oil” or “bio-crude”, it has few similarities with petroleum. Unlike the hydrocarbons found in petroleum, it contains very high levels of oxygen, comparable to that of the biomass from which it is derived. On a mass basis, its heating value is comparable to biomass. In fact, it is sometimes characterized as “liquid biomass”. Bio-oil is composed of a complex mixture of water, volatile oxygenates, anhydrosugars, and non-volatile oligomers. Many of these compounds are extremely reactive, leading to polymerization of bio-oil even at ambient temperatures. These physiochemical properties make bio-oil unsuitable as transportation fuel without upgrading. A number of upgrading processes have been explored, including catalytic cracking, hydroprocessing, aqueous phase reforming, and fermentation, to produce transportation fuels from fast pyrolysis process.²¹ 

The pace of research and development in fast pyrolysis has accelerated considerably in the past decade. As shown in Figure 1.3, scientific publications increased from a dozen or so per year in the early 1980s to over 700 per year in 2016. Most of this growth occurred in the last decade in a period when development of advanced biofuels became a priority in the United States and many other countries around the world.

This rapid growth in research and development has not yet translated into notable successes in the commercial deployment of fast pyrolysis technologies.²²  A number of factors have contributed to the slow pace of commercialization. The meteoric rise in petroleum prices in the first decade of the twenty first century was followed in the next decade by a dramatic price slump, hurting the economics of advanced biofuels. Failure of many markets to assign a cost to carbon emissions from fossil fuels has also diminished the incentive for the commercialization of low carbon fuels. However, in some respects, commercialization efforts launched by the U.S. Energy Security Act of 2005 got ahead of the state-of-the art technology in advanced biofuels: crucial scientific questions and engineering challenges related to pyrolysis as well as other advanced biofuels technologies had not been resolved. Much has changed in the past decade. This book covers recent advances in fast pyrolysis science and technology, with chapters on reaction chemistry of thermal deconstruction, computational modeling, product utilization, catalytic pyrolysis and economic evaluation of fast pyrolysis technology.

As the most abundant polymer on the planet and the major component of lignocellulosic biomass, the pyrolysis of cellulose has been studied since the early 20th century.²³  Although it is generally accepted that levoglucosan is the primary product of cellulose fast pyrolysis,¹³  the mechanism by which it forms and its role in producing secondary products continues to be debated. It has been reported recently that cellulose pyrolysis proceeds through the formation of a liquid intermediate, from which volatile products are formed.²⁴  Multiphase reactions are involved in cellulose pyrolysis, including solid–liquid, liquid–liquid, liquid–gas, and gas–gas reactions. Moreover, reaction intermediates may exist for only fractions of a second,²⁴  further complicating experiment investigations of cellulose pyrolysis. Chapter 2 is a historical review of the many studies of cellulose pyrolysis covering the several theories and unresolved questions around the primary reactions of cellulose depolymerization.

As described in Chapter 3, lignin is the most complex and difficult to pyrolyze component of lignocellulosic biomass. Lignin melts before it depolymerizes. The liquid products of lignin pyrolysis include volatile phenolic monomers and non-volatile phenolic oligomers with molecular weights ranging up to 2500 Da. These phenolic compounds are extremely reactive, often condensing and dehydrating to char and light gas before they can volatilize. These characteristics of lignin can cause agglomeration in reactors and clogging of condensers.

Due to the inherent reactivity of pyrolysis products, various secondary reactions could occur during biomass pyrolysis, leading to a lower yield of bio-oil with undesirable properties. Chapter 4 reviews recent studies related to secondary reactions of pyrolysis of cellulose, lignin, and whole biomass. Heat and mass transfer phenomena during fast pyrolysis are also addressed in this chapter.

The role of free radicals in pyrolysis has been the subject of speculation for many years. The short half-life of radicals is among the challenges in studying them. Recent advances in analytical techniques such as in situ Electron Paramagnetic Resonance have been employed to probe free radical chemistry during pyrolysis. Recent studies^25,26  have revealed that free radicals participate in a variety of reactions, especially the depolymerization of lignin. Chapter 7 reviews studies that have detected free radicals in bio-oil and biochar and investigated their role in fast pyrolysis chemistry. The role of radicals in bio-oil stability is also discussed.

Improved understanding of the physics and chemistry of fast pyrolysis arising from experimental studies has encouraged multi-scale modeling of pyrolysis processes. Intra-particle chemistry and transport processes are receiving particular attention from the scientific community as their role in determining overall yields and selectivity of pyrolysis products becomes increasingly evident. Chapter 11 describes recent advances in modeling pyrolysis in individual particles and provides readers the state-of-the-art in predicting pyrolysis kinetics. Limitation of current modeling efforts and future opportunities are also discussed.

Fast pyrolysis proceeds through a complex set of primary and secondary reactions leading to a large number of organic compounds with a wide range of properties, making characterization of bio-oil challenging. Detailed characterization of bio-oil is valuable for both interpreting pyrolysis reaction chemistry and advancing its upgrading to transportation fuels or refining to biobased chemicals. Chapter 6 covers the physiochemical properties of bio-oil and the corresponding characterization methods.

One of the most distinguishing characteristics of bio-oil is its high reactivity compared to petroleum, which presents one of the chief challenges in using it as feedstock in petroleum refineries. Unlike the hydrocarbons in petroleum, bio-oil is highly oxygenated and includes alcohols, aldehydes, ketones, carboxylic acids, furans, sugars, and phenolic compounds. These compounds can participate in a wide variety of reactions contributing to a phenomena referred to as “instability” or “aging”, characterized by an increase in viscosity, molecular weight, and water content of the bio-oil and ultimately phase separation.²⁷  In ambient conditions, instability can manifest itself over time periods as short as a few weeks. At the elevated temperatures employed for upgrading bio-oil, this instability can occur within minutes. Stability problems need to be solved to produce bio-oil with uniform characteristics in commercial scale. Chapter 8 reviews the complex phenomenon of bio-oil instability and summarizes stabilization methods including dilution, neutralization, esterification, mild hydrotreating, and stage-fractionation.

The production of chemicals has received significant attention as a way to valorize bio-oil. However, its complex chemical composition and reactivity makes the separation of bio-oil molecules by functional groups challenging.⁹  Chapter 9 provides an overview of relevant separation and enrichment techniques for bio-oil, including distillation, liquid–liquid extraction, adsorption/desorption, and fractional condensation. A specific example of chemical recovery from bio-oil at an industrial scale is also briefly reported.

Although fast pyrolysis bio-oil has prospects for being upgraded into hydrocarbon fuels, raw bio-oil is presently incompatible with processing at traditional petroleum refineries. One of the challenges is that the elemental composition of bio-oil more closely resembles biomass than petroleum. Some of the chemical and physical properties of crude bio-oil that prevent its processing in petroleum refineries include its high oxygen content and acidity and poor thermal stability. In an effort to improve the quality of bio-oil for upgrading, recent research has focused on catalytically upgrading pyrolysis vapors before they are condensed to bio-oil. Known as catalytic fast pyrolysis, the goal is to substantially, if not completely, deoxygenate the products, producing molecules that are suitable as blendstock for refining to hydrocarbon transportation fuels.²⁸  Among the catalysts available, inexpensive zeolites are particularly attractive due to their ability to deoxygenate bio-oil molecules without the addition of hydrogen.²⁹  Chapter 10 explores the ability of zeolite catalysts to produce a stable, upgradable product from bio-oil.

Low carbon efficiency and excessive coke formation on catalysts are two of the hurdles in the further development of catalytic fast pyrolysis. Due to the hydrogen-deficient nature of biomass, the liquid product from catalytic pyrolysis can become even more hydrogen deficient as dehydration occurs. In order to improve the carbon efficiency and bio-oil quality, the addition of reactive gases, such as hydrogen and carbon monoxide, to the pyrolysis process, with or without deoxygenation catalysts, has been proposed. Very recently, several fast pyrolysis technologies involving reactive gases have been developed.^30–33  One of these novel processes is tail gas recycling pyrolysis (TGRP), which recycles the pyrolysis gases for fluidization.³²  Deoxygenated bio-oil with improved quality was observed for TGRP in a reducing environment. However, the detailed mechanism of bio-oil deoxygenation via tail gas recycling is still unclear. Another example is the IH² process developed by the Gas Technology Institute,^30,31  which integrates hydropyrolysis of biomass with close-coupled hydroprocessing of the hydropyrolysis vapors to produce gasoline and diesel. Chapter 5 provides an overview of the recent studies in the role of reactive gases in improving the products of fast pyrolysis.

There are no successful commercial projects from which to estimate the cost of fuels and chemicals from fast pyrolysis. Techno-economic analyses based on limited yield data and well-established process engineering principles have been conducted to estimate the economic feasibility of fast pyrolysis of biomass in comparison to other fuel and chemical pathways including biochemical processing of biomass and conventional petroleum refining.^34,35  Chapter 12 reviews recent economic assessments of the major pyrolysis pathways for advanced biofuel production. Suggestions are also made to improve the accuracy of existing deterministic techno-economic models.