Introduction

The invention relates to a process for the controlled conversion of particulate matter comprising hemicellulose, cellulose and lignin, more specifically it relates to a controlled process for the conversion of a particulate lignocellulosic material.

Background of the invention

Currently, many chemical building blocks and synthetic carbon-containing products used in chemical industry as a source for further products like plastics are based on fossile carbon sources (e.g. mineral oil, gas). For reasons of limited (future) supply of such fossile sources and the impact of usage of such sources on greenhouse gasses and their (potential) impact on climate change, there is a demand for production of carbon-containing chemical building blocks and base chemicals from renewable sources. At the same time, it is desired that such carbon-containing chemical building blocks and base chemicals are not based on material which is (or can easily be made) edible for humans, as such would take an undesired demand on arable land. For these reasons, lignocellulosic materials have been studied at for the purpose of serving as a source for renewable chemicals. Lignocellulosic material is generally biomass not or less suitable for human consumption, which biomass usually contains as main constituents cellulose, hemicellulose and lignin. Examples of such lignocellulosic material include wood, straw, nutshells, corn stover and bagasse.

Several processes have been studied in the past to obtain useful materials from such lignocellulosic material. An example of such is the Bergius-Rheinau process. In the Bergius-Rheinau process solid lignocellulosic material, such as wood, is treated with a concentrated hydrochloric acid composition. Such hydrochloric acid treatment may result in (partial) hydrolysis of the cellulose and hemicellulose and thus give a hydrolysate and a residue that consists for a large part of lignin. From the hydrolysate of cellulose and hemicellulose saccharides (typically mono- and oligosaccharides) may be obtained, which saccharides can be used in further conversion processed to make e.g. ethanol, ethyleneglycol and other (base) chemicals. This Bergius-Rheinau process has been described by F. Bergius, Current Science , Vol. 5, No. 12 (June 1937), pp. 632-637 and the hydrolysis step is in essence a one-stage hydrolysis using hydrochloric acid of 40%, which hydrolyses both hemicellulose and cellulose. A hydrolysate obtained with such process contains both saccharides originating from hemicellulose (e.g. xylose, arabinose, mannose, glucose and their oligomers) and cellulose (mainly glucose and its oligomers). GB827921 discloses a process for producing sugars from cellulosic materials containing cellulose, lignin and hemicellulose, by contacting such cellulosic material with concentrated hydrochloric acid, and obtaining the hydrolysate of hydrolysed hemicellulose and optionally hydrolysed cellulose.

The hydrolysis of the hemicellulose and cellulose may also be effected in two-stages: a first stage hydrolyzing mainly hemicellulose and a second stage hydrolyzing mainly cellulose. The advantage of such is that the resulting saccharide fractions can be obtained separately, which provides more options for adding value to the resulting hydrolysates. An example of such a two-stage hydrolysis of lignocellulosic biomass is described in US 294577, which is a Bergius-Rheinau type process modified by the patentee (Riehm). The process disclosed therein uses a hydrochloric acid solution of 34-37% for hydrolyzing the hemicellulose fraction of the lignocellulosic biomass (e.g. pinewood sawdust) first (named prehydrolysis) followed by a hydrolysis of the cellulose fraction of the remaining material using a hydrochloric acid solution of 40-42% (named main hydrolysis).

A more recent example of such two-stage hydrolysis is disclosed in WO2016/082816. In the process in this reference in a vertical reactor filled with vegetable biomass particle, hydrochloric acid of 35-37% is fed from below into the reactor to effect hydrolysis of the hemicellulose, followed by feeding at the bottom of the reactor a 40-42% hydrochloric acid solution (to effect hydrolysis of cellulose) which displaces the 35- 37% hydrochloric acid solution. It is stated that the flowspeed of the hydrochloric acid should be such that displacement of the lower concentrated acid by the higher concentrated acid would lead to minimal mixing of both acid fractions, without giving any further indication as to how this needs to be effected.

There is a need for a Bergius Rheinau-type process (i.e. hydrolysis of saccharides in lignocellulosic biomass using strong hydrochloric acid) wherein the pre-hydrolysate (containing saccharides mainly resulting from hydrolysis of hemicellulose) and main hydrolysate (containing saccharides manly resulting from hydrolysis of cellulose) that are formed can be obtained largely separated from one another in an easy way, as such may lead to more pure saccharide streams, such as glucose and its oligomers (and thus a potential for more value creation). Preferably, such process should be easy to control, be robust and not overly complex, time efficient, and yields (amount of hemicellulose and cellulose that can be converted into saccharides and its oligomers and obtained) should preferably high. The process should preferably also allow for different particulate lignocellulosic biomass sources, with different compositions. The process should preferably not consume more hydrochloric acid than necessary for hydrolyzing the cellulose and hemicellulose to a desired degree. Hence, the process should allow for relatively straightforward process control, e.g. in terms of the amounts of acids required and/or duration of hydrolysis. Preferably, the process be economical on the amount of hydrochloric acid for pre- or main hydrolysis required, and/or should facilitate performing the process in optimum time.

Summary of the invention

It has now been found that the above objective can be met, at least in part, by a process for obtaining hydrolysis products from particulate matter comprising hemicellulose, cellulose and lignin, by subjecting a reactor comprising said particulate matter and interstitial space to the following steps:

a. feeding to said reactor a mass "Ml" of a first hydrochloric acid solution to hydrolyse at least part of the hemicellulose of said particulate matter by contacting the particulate matter for a duration "tl" with said first aqueous hydrochloric acid solution, which first aqueous hydrochloric acid solution has a hydrochloric acid concentration of between 30 wt. % and 42 wt. %, based on the weight amount of water and hydrochloric acid in such first aqueous hydrochloric acid solution, yielding a first remaining particulate matter and a first aqueous hydrolysate product solution; b. feeding to said reactor a water-immiscible displacement fluid in a volume DF1 thereby displacing at least part of said first aqueous hydrolysate product solution from the interstitial space with said water-immiscible displacement fluid;

c. feeding to said reactor a mass "M2" a second hydrochloric acid solution to hydrolyse at least part of the cellulose of the first remaining particulate matter by contacting the first remaining particulate matter for a duration "t2" with said second hydrochloric acid solution, which second hydrochloric acid solution has a hydrochloric concentration of between 40% and 51%, based on the weight amount of water and hydrochloric acid in the second hydrochloric acid solution whilst said second hydrochloric acid solution has a hydrochloric acid concentration which is the same or higher than the first hydrochloric acid solution added in step a., yielding a second remaining particulate material and a second aqueous hydrolysate product solution;

wherein mass "Ml" in step a. and/or mass "M2" in step c. and optionally duration tl and/or t2 are estimated prior to the feeding of step a. by one or more of:

A. an analysis of the particulate matter comprising hemicellulose, cellulose and lignin;

B. a trial hydrolysis experiment of the particulate matter comprising hemicellulose, cellulose and lignin using hydrochloric acid; C. an estimation of the theoretical amount of the first hydrochloric acid solution that is needed for a desired degree of hydrolysis of the hemicellulose:

D. a mathematical model.

Previously, a process has been developed as set out in WO2019149833, wherein the pre-hydrolysis (of hemicellulose) and main hydrolysis (of cellulose, using hydrochloric acid of greater concentration than for the pre-hydrolysis) are separated by using a displacement fluid. In the process of said reference, all three liquids (hydrochloric acid for pre-hydrolysis, displacement fluid, and hydrochloric acid for main-hydrolysis) flow through a reactor which contains lignocellulosic biomass particles one after the other. If managed well (w.r.t. where to pump in which fluid and where to leave out fluid form the reactor, e.g. top or bottom) first the hydrochloric acid used for pre-hydrolysis (and thus pre-hydrolysate obtained) flows through the reactor bed, at some point pushed out by displacement fluid, in turn pushed out by acid used for main hydrolysis (and thus main hydrolysate obtained), optionally followed by further displacement fluid. Such process works well for a reactor which contains the lignocellulosic biomass in particulate form, leaving some interstitial space between the particulates. This process is now adapted with a series of steps that lead to an informed estimation on the amounts of hydrochloric acid that is required for an optimal hydrolysis process. The advantage of such is that it is easier to use different sources of lignocellulosic biomass, and tailoring more specifically the amount of hydrochloric acid needed for such. Said estimation can also be used for determining or estimating the optimal duration of both hydrolysis steps (i.e. how long the particulate matter should be subjected to the acid treatment). Hence, in the present invention it is preferred that duration "ti" and/or "t ₂ " in the process are estimated prior to the feeding of step a. by an analysis of the particulate matter and/or a mathematical model.

Detailed description of the invention

In the present invention, "particulate matter" herein in connection with the lignocellulosic biomass refers to material which is not liquid or gaseous but solid, and which is at the same time divided up in units such than when the reactor is filled with the particulate matter a bed is obtained which also contains interstitial space through which fluids can flow. For clarity, "particulate matter" herein covers fairly hard pieces such as woodchips and pieces of coconut shell but also fibrous material such as bagasse and particles made out if such. "Interstitial space" herein means the voids in a reactor filled with particulate matter, or in other words the space inside the reactor but outside the particulate matter.

"Water-immiscible" herein means, in connection to the displacement fluid and displacement liquid, that such displacement fluid or displacement liquid has a solubility in water of less than 3 g displacement fluid (or displacement liquid) per litre of water, at 20°C and atmospheric pressure. Preferably, such solubility is less than 2 g/L, even more preferably less than 1 g/L, under such conditions.

To a large extent, the amounts Mi and M ₂ of hydrochloric acid needed for the process for obtaining hydrolysis products from particulate matter comprising hemicellulose, cellulose and lignin using hydrochloric acid in the type of process concerned depend on the composition of the particulate matter used as starting material. Clearly, if particulate matter is used as starting material that has a relatively high amount of hemicellulose, it can be expected that more acid (or longer time) is needed for step a. in the present process (the pre-hydrolysis). However, the reality is that also the more specific nature of the hemicellulose and how it is integrated in the particulate matter used as a starting material plays a role. Likewise, the size and shape of the particulate matter also may be of influence on how easy or fast hydrolysis of hemicellulose or cellulose goes, how much acid is needed and how long the acid needs to be allowed to react on the particulate biomass. Additionally, the nature of how the particulate matter is packed in the reactor (e.g. how much interstitial space or conversely bed density) may be of influence on the speed and extent of hydrolysis and thus the amount of acid needed and/or the time required for the acid to act on the particulate matter. For that reason, it is preferred that mass "Mi" in step a. and/or mass "M ₂ " in step c. and optionally duration "ti" and/or "t ₂ " are estimated prior to the feeding of step a. taking into account such factors on particulate matter. This is achieved by the present invention.

In the process according to the present invention, it is preferred that the analysis of the particulate matter comprising hemicellulose, cellulose and lignin in A comprises one or more of: a determination of the amount of hemicellulose in the particulate matter to be subjected to hydrolysis, the amount of cellulose in the particulate matter to be subjected to hydrolysis, the particle size distribution of the particulate matter to be subjected to hydrolysis, a bed density of a packed bed of particulate matter to be subjected to hydrolysis. In the above, the amount of hemicellulose and/or cellulose in the particulate matter to be hydrolysed may also be estimated, e.g. based on comparison with values found in literature or obtained from suppliers or analysis of comparable matter. It will be clear that it is most beneficial for the purpose of this invention if any trial hydrolysis experiment reflects some of the conditions of the actual hydrolysis which is aimed at production of saccharides.

Hence, in the present invention it is preferred that the trial hydrolysis experiment of the particulate matter comprising hemicellulose, cellulose and lignin using hydrochloric acid is conducted for a time of between 5 and 50 hours in at least one single flow-through trial reactor, wherein the flow in such reactor comprises hydrochloric acid having a concentration which is between 90% and 110% of the concentration of the first hydrochloric acid solution , and wherein of the flow out the concentration of hydrolysis products of at least the hemicellulose is measured.

In the process as set out above it will be clear that if certain compositional facts are known, a more accurate estimation can be made of the theoretical amount of acid needed. Hence, in the above process, it is preferred that the estimation in C is based on the analysis results of A and/or B. In the present invention the mathematical model in D. is preferably a mass balance model.

In a preferred way, in the process of the present invention, mass "Mi" in step a. and/or mass "M ₂ " in step c. and optionally duration ti and/or t ₂ are estimated prior to the feeding of step a. by the combination of:

A. an analysis of the particulate matter comprising hemicellulose, cellulose and lignin;

B. a trial hydrolysis experiment for a time of between 5 and 50 hours of the particulate matter in at least one single flow-through trial reactor, wherein the flow in such reactor comprises hydrochloric acid having a concentration which is between 90% and 110% of the concentration of the first hydrochloric acid solution, and wherein of the flow out the concentration of hydrolysis products of at least the hemicellulose is measured;

C. an estimation of the theoretical amount of the first hydrochloric acid solution and the second hydrochloric acid solution that is needed for a desired degree of hydrolysis of the hemicellulose and cellulose;

D. using a mass balance model to estimate Mi and/or M ₂ and optionally ti and/or t ₂ , using as input for such model one or more of: the outcome of B, the outcome of C, the number of reactors, the reactor volume, the amount of particulate matter comprising hemicellulose, cellulose and lignin in the process.

By following the above analyses and estimations and experiments in combination, preferably sequential in the order given, information is obtained both of the nature of the particulate matter itself as well as its behavior in a (simple) single hydrolysis process, which information can be used to make a better estimate of the amounts of hydrochloric acid needed and optionally the time involved for the hydrolysis steps.

As to the above process, it is preferred that the analysis in A. comprises one or more of: a determination of the amount of hemicellulose in the particulate matter to be subjected to hydrolysis, the amount of cellulose in the particulate matter to be subjected to hydrolysis, the particle size distribution of the particulate matter to be subjected to hydrolysis, a bed density of a packed bed of particulate matter to be subjected to hydrolysis. Regarding this, it will be clear that biomass sources will differ in the amounts of hemicellulose and cellulose and that this has an effect of the amount of hydrochloric acid required.

Likewise, it will be clear that if the particulate matter is composed of large particles, more acid is expected to be needed, as there may be a diffusion limitation. Additionally, the bed density (of particulates to be hydrolysed in a reactor) may determine how (fast or slow) a hydrochloric acid solution flows through a bed in a reactor containing such particles.

In the process as set out above it will be clear that if certain compositional facts are known, a more accurate estimation can be made of the theoretical amount of acid needed. Hence, in the above process, it is preferred that the estimation in C is based on the analysis results of A and/or B.

As to step B. in the above process, it is preferred that such is representative of the hydrolysis process as in the end will be chosen for production, e.g. in terms of reaction time and reactor size, for the best estimation. Hence, in such process it is preferred that in the hydrolysis trial of step B. the hydrolysis is carried out for a time of between 10 and 36 hours, and preferably wherein the single flow-through trial has a volume which is at least 30% of the size of the volume of the hydrolysis reactor.

The mass balance model in step D. in the above process can easily be carried out using a spreadsheet, which is fed with data. Optionally, such data may also include results of actual full hydrolysis processes that have been carried out (i.e. steps a. to c.). Apart from giving an estimation on the amounts of hydrochloric acid solution and optionally time for hydrolysis, this process may also be used to estimate the amount of displacement fluid that is required in the present process, in particular for the displacement fluid in step b. of the present process, but also any displacement fluid that is used optionally as a step d., after step c. This is in particular so as such follows fairly directly from a mass balance: the amount of hemicellulose and cellulose are known, the amount of acid that goes in, the amount of hydrolysate going out, and the volume reduction in the particulates that occurs due to hemicellulose and/or cellulose becoming hydrolysed and dissolved. Hence, it is furthermore preferred that in the above process the mass balance model in step D is also used to estimate the volume DFi of water-immiscible displacement fluid used in step b., and also optionally of the volume of displacement fluid DF ₂ , if such displacement fluid is used as an additional step d. after c.

In the process according to the present invention, it is desired that the prior analysis or prior hydrolysis gives an indication for the amount of hydrochloric acid to be used. Hence, in the present invention, it is preferred that mass "Mi" and mass "M ₂ " are expressed as weight of hydrochloric acid solution per weight of particulate matter.

For the process according to the present invention, it is preferred that the first hydrochloric acid solution has a concentration of between 33 and 40 wt. %, based on the weight amount of water and hydrochloric acid in such first aqueous hydrochloric acid solution. More preferably such concentration is between 35 and 38 wt%, based on the weight amount of water and hydrochloric acid in such first aqueous hydrochloric acid solution. The concentration of the second hydrochloric acid solution used in the process according to the present invention is preferably between 40 and 46 wt. %, based on the weight amount of water and hydrochloric acid in such second aqueous hydrochloric acid solution, more preferably between 40 and 44 wt. %, based on the weight amount of water and hydrochloric acid in such second aqueous hydrochloric acid solution. However, the concentration of the second hydrochloric acid solution should be higher than that of the first. Hence, the lower range (e.g. 40-42%) of concentration given for the second hydrochloric acid can only be applied if the concentration of the first hydrochloric acid has a concentration of e.g. between 30 and 39 wt%, more likely 30-37 wt%. As already indicated by Bergius (publication under Background of the Invention), an advantage of hydrolysis using strong hydrochloric acid is that it can be carried out at ambient temperature and pressure. Hence, in the present invention it is preferred that the first hydrochloric acid and second hydrochloric acid added in steps a. and c. to the reactor are at a temperature of between 1 and 40°C, preferably between 5 and 30°C, and that the pressure in the reactors during steps a-c is about 0.1 MPa (atmospheric pressure).

During step (a) hemicellulose is being hydrolyzed and the resulting saccharides (typically a mixture of mono-, di-, and oligosaccharides) become dissolved in the first aqueous hydrochloric acid solution.

Therefore, in addition to the water and the hydrochloric acid, the first aqueous hydrochloric acid solution may or may not contain other compounds such as for example dissolved saccharides. Similarly, during step (c) cellulose is being hydrolyzed and the resulting saccharides (typically a mixture of mono-, di-, and oligosaccharides) become dissolved in the second aqueous hydrochloric acid solution. Therefore, in addition to the water and the hydrochloric acid, the second aqueous hydrochloric acid solution may or may not contain other compounds such as for example dissolved saccharides. The process of subsequent steps a-c (and optionally d) may be carried out in one or more reactors. Preferably, the process is carried out in at least two reactors in series wherein the reactors are at different stages in the process sequence of a-c (or d). Also, multiple reactors may be used for step a and also for step c (and if desired also for step b or d, although such is less logical).

Suitable processes for obtaining a saccharide product from the pre-hydrolysate solution (i.e. the first hydrolysate product solution) and/or the main hydrolysate solution (i.e. the second hydrolysate solution) are described in for example WO2017/082723 and WO2016/099272. Preferably the pre-hydrolysate solution and/or the main hydrolysate solution is suitably first admixed with a carrier liquid, in which the saccharides are insoluble and that has a boiling point higher than that of water to obtain an aqueous admixture. Subsequently such aqueous admixture can be subjected to an evaporation step, to yield a vapor fraction comprising water and hydrochloric acid and a residue fraction comprising solid saccharides and the carrier liquid. The vapor fraction may advantageously be condensed, reconcentrated and recycled to the process to be used as a first or second hydrochloric acid solution. The residue fraction comprising solid saccharides and the carrier liquid can conveniently be recovered and passed to a separation vessel. Such a separation vessel can for example be a settling vessel or any other separator that is suitable to separate the saccharides from the carrier liquid. From the separation vessel a saccharide product can be obtained. In addition a stream of crude carrier liquid can be obtained that can be cleaned and recycled. Thus, preferably the process according to invention comprises one or more further steps wherein:

the first hydrolysate product solution and/or the second hydrolysate product solution is/are admixed with a carrier liquid, in which saccharides are insoluble and that has a boiling point higher than that of water to obtain an aqueous admixture;

the aqueous admixture is subjected to an evaporation step, to yield a vapor fraction comprising water and hydrochloric acid and a residue fraction comprising solid saccharides and the carrier liquid; and

the residue fraction comprising solid saccharides and the carrier liquid is passed to a separation vessel to obtain a saccharides product.

As mentioned herein before, a process has been developed as set out in PCT/EP2019/052404, wherein the pre-hydrolysis (of mainly hemicellulose) and main hydrolysis (of mainly cellulose, using hydrochloric acid of greater concentration than for the pre-hydrolysis) are separated by using a displacement fluid. In the process of said reference, all three liquids (hydrochloric acid for pre-hydrolysis, displacement fluid, and hydrochloric acid for main-hydrolysis) flow through a reactor one after the other, which reactor contains lignocellulosic (biomass) particles. As stated under Summary of the invention, following step b. the water- immiscible displacement fluid displaces at least part of the first aqueous hydrolysate product solution obtained by step a. from the interstitial space with said water-immiscible displacement fluid. Similarly, the feeding to the reactor of said second hydrochloric acid solution in step c. may displace (and this is preferred) at least part of the water-immiscible displacement fluid from step b., thereby effecting removal of at least part of said water-immiscible displacement fluid from the interstitial space. Hence, it may be preferred that an additional step d. is added to the present process, of feeding to said reactor a water- immiscible displacement fluid thereby displacing at least part of said second aqueous hydrolysate product solution from the interstitial space with said water-immiscible displacement fluid. The water-immiscible displacement fluid used for such additional step d. may use a different water-immiscible displacement fluid or the same as was used for step b. It is preferred that these are the same. Additionally, it can be convenient to re-use the water-immiscible displacement fluid. In such a case, water-immiscible displacement fluid can be retrieved from step (c) and recycled to step (b). The water-immiscible displacement fluid retrieved from step (c) can optionally be purified and/or can optionally be stored in a displacement fluid storage vessel before being recycled to step (b). If there is an additional step d. in which water-immiscible displacement fluid displaces at least part of the second aqueous hydrolysate product solution the same applies: it may rely on recycled displacement fluid.

As to the displacement fluid, it is preferred that it is a water-immiscible liquid (water-immiscibility as defined above). More preferably, the displacement fluid in the present process is a water-immiscible displacement liquid having a boiling temperature at 0.1 MPa of equal to or more than 50°C, more preferably equal to or more than 80°C and even more preferably equal to or more than 100°C. Preferably, the water-immiscible displacement fluid has a melting temperature at ambient pressure (i.e. at 0.1 MegaPascal) of equal to or less than 0°C, more preferably equal to or less than minus 5 degrees Celsius (- 5°C), even more preferably equal to or less than minus 10 degrees Celsius (-10°C) and still more preferably equal to or less than minus 20 degrees Celsius (-20°C). Preferably, the water-immiscible displacement fluid has no flash point or a flash point equal to or more than 60°C, even more preferably equal to or more than 80°C and still more preferably equal to or more than 100°C. Such a flashpoint may for example be determined by ASTM method no. ASTM D93. Clearly, for the displacement liquid to easily flow through the interstitial space of the reactor, it is preferred that the viscosity is not unduly high. Hence, it is preferred that the water-immiscible displacement liquid has a viscosity at 20°C of equal to or less than 5 centipoise (cP), more preferably equal to or less than 4.0 cP and most preferably equal to or less than 2 cP. Such viscosity may for example be determined by ASTM method no. ASTM D445 - 17a. Additionally, it is preferred that the water-immiscible displacement fluid is a liquid having a density equal to or less than 1200 kilograms per cubic meter (kg/m ³ ), even more preferable a liquid having a density equal to or less than 1000 kg/m ³ and still more preferably a liquid having a density equal to or less than 800 kg/m ³ . Such density may for example be determined by ASTM method no. ASTM D1217 - 15. Preferably, the displacement fluid is essentially water-free, and preferably essentially immiscible with an aqueous hydrochloric acid solution and/or an aqueous first hydrolysate product solution and/or an aqueous second hydrolysate product solution as described herein. Preferably, the water-immiscible displacement liquid comprises or consists of one or more alkanes, more preferably one or more alkanes having in the range from equal to or more than 5 to equal to or less than 20 carbon atoms, even more preferably an alkane having in the range from equal to or more than 6 to equal to or less than 16 carbon atoms. The alkanes may be cyclic or non-cyclic. Most preferably, the water- immiscible displacement liquid comprises or consists of one or more alkanes chosen from the group consisting of cyclic hexane, normal hexane, iso-hexane and other hexanes, normal heptane, iso-heptane and other heptanes, normal octane, iso-octane and other octanes, normal nonane, iso-nonane and other nonanes, normal decane, iso-decane and other decanes, normal undecane, iso-undecane and other undecanes, normal dodecane, iso-dodecane and other dodecanes, normal tridecane, iso-tridecane and other tridecanes, normal tetradecane, iso-tetradecane and other tetradecanes, normal pentadecane, iso- pentadecane and other pentadecanes, normal hexadecane, iso-hexadecane and other hexadecanes.

The processes of the present invention will work well if in a reactor packed with particulate matter if there is still some interstitial space, through which the hydrochloric acid and displacement fluid can percolate. For such, in the present invention it is preferred that the reactor comprising said particulate matter and interstitial space has a porosity calculated as of between 0.1 and 0.5, preferably said porosity is between 0.2 and 0.4, wherein and is the volume in such.

Typically, in the process according to the present invention the particulate matter comprising

hemicellulose, cellulose and lignin is preferably particulate matter of vegetable biomass. The particulate matter may conveniently be washed, dried, roasted, torrefied and/or reduced in particle size before it is used as a feedstock in the process according to the invention. The particulate matter may conveniently be supplied or be present in a variety of forms, including chips, pellets, powder, chunks, briquettes, crushed particles, milled particles, ground particles or a combination of two or more of these. Suitable examples of such particulate matter include wood chips, preferably woodchips from softwood or rubberwood.

Examples

Example 1

Non-limiting figures 1A, IB, 1C, 2A and 2B illustrate an example of a process of hydrolysing particulate matter containing hemicellulose, cellulose, and lignin, with hydrochloric acid. A brief description of the figures of this example:

Figures 1A, IB and 1C illustrate a first cycle, starting at a time "t", of a process according to the invention. Figures 2A and 2B illustrate a second subsequent cycle, starting at a time "t+8 hours", of the same process as figures 1A, IB and 1C.

The illustrated process is carried out in a reactor sequence of 6 hydrolysis reactors (R1 to R6). The hydrolysis reactors are operated at a temperature of 20°C and a pressure of 0.1 MegaPascal. The process is operated in a sequence of cycles, each cycle being carried out within a 8 hour cycle period.

[0001 ] Figure 1A illustrates the start of a new cycle. At the start of a new cycle, dried wood chips (101) have just been loaded into reactor (Rl) via solid inlet line (102). Reactor (R2) contains an intermediate prehydrolysate solution and a solid material containing cellulose and lignin. The hemicellulose is already at least partly hydrolysed. Reactor (R3) contains a displacement fluid (such as for example iso-octane) and a solid material containing cellulose and lignin. Reactors (R4) and (R5) each contain an intermediate hydrolysate solution. The intermediate hydrolysate solution in reactor (R4) can contain a higher amount of saccharides than the intermediate hydrolysate solution in reactor (R5), as explained below. In addition reactors (R4) and (R5) contain a solid material containing lignin. The cellulose is already at least partly hydrolysed. Reactor (R6) contains a displacement fluid (such as for example iso-octane) and a residue. The residue is a solid material containing lignin.

As illustrated in figure IB, during a first part of the cycle, reactor (Rl) is flooded with a plug (104c) of intermediate prehydrolysate solution coming from a storage vessel (103), a plug (104a) of fresh first aqueous hydrochloric acid solution is introduced to reactor (R2), a plug (105a) of fresh second aqueous hydrochloric acid solution is introduced to reactor (R5) and a plug (106d) of displacement fluid is drained from reactor (R6). After reactor (Rl) has been flooded with a plug (104c when going into Rl, 104d when being pushed out of Rl) of intermediate prehydrolysate solution coming from a storage vessel (103), a plug (104a) of fresh first aqueous hydrochloric acid solution, having a hydrochloric acid concentration of 37.0 wt. % and containing essentially no saccharides yet, is introduced into reactor (R2), thereby pushing forward a plug (104b) of intermediate pre-hydrolysate solution, containing hydrochloric acid in a concentration of about 37.0 wt.

%, but also containing already some saccharides (i.e. saccharides derived from solid material that was residing in reactor (R2)), from reactor (R2) into reactor (Rl).The plug (104b) of intermediate pre hydrolysate solution, pushes the plug (104d) out from reactor (Rl). Plug (104d) previously contained intermediate pre-hydrolysate solution, but has now taken up sufficient saccharides and has become a final first hydrolysate product solution. Such final first hydrolysate product solution can suitably be forwarded to one or more subsequent processes or devices, where optionally hydrochloric acid could be removed from the pre-hydrolysate solution and recycled.

During the same first part of the cycle, a plug (105a) of fresh second aqueous hydrochloric acid solution, having a hydrochloric acid concentration of 42.0 wt. % and containing essentially no saccharides yet, is introduced into reactor (R5), thereby pushing forward a plug (105b) of intermediate hydrolysate solution, containing hydrochloric acid in a concentration of about 42.0 wt. %, but also containing already some saccharides (i.e. derived from the solid material that was residing in reactor (R5)), from reactor (R5) into reactor (R4). This plug (105b) in its turn pushes forward a second plug (105c) of intermediate hydrolysate solution, containing hydrochloric acid in a concentration of about 42.0 wt. %, but also containing saccharides (i.e. derived from solid material that was residing in previous reactors), from reactor (R4) into reactor (R3). Whilst being pushed from reactor (R5) into reactor (R4) and further into reactor (R3), the intermediate hydrolysate solution absorbs more and more saccharides from the solid material remaining in such reactors from previous stages. The saccharide concentration of the intermediate hydrolysate solution advantageously increases, thus allowing a saccharide concentration to be obtained, that is higher than the saccharide concentration obtained in a batch-process.

The plug (105c) of intermediate hydrolysate solution being pushed from reactor (R4) into reactor (R3), pushes a plug (106c) of displacement fluid out of reactor (R3).

During this same first part of the cycle, further a plug (106d) of displacement fluid is drained from reactor (R6), leaving behind a residue containing lignin. During a second part of the cycle, as illustrated by figure 1C, a plug (106a) of displacement fluid is introduced into reactor (R2). This plug (106a) may or may not contain parts of the plug (106c) of displacement fluid that was pushed out of reactor (R3). Advantageously, the volume of displacement fluid in plug (106a) can be adjusted, for example by adding more or less displacement fluid, to compensate for volume losses due to the reduction of solid material volume. This allows one to ensure that all reactors remain sufficiently filled with volume and it allows one to maintain a sufficient flowrate.

The plug (106a) of displacement fluid being introduced in reactor (R2), suitably pushes forward plug (104a) that was residing in reactor (R2). Plug (104a), previously contained merely fresh first aqueous hydrochloric acid solution, but has in the meantime taken up saccharides from the solid material in reactor (R2) and has become an intermediate pre-hydrolysate solution. Plug (104a) is pushed out of reactor (R2) into reactor (Rl), thereby pushing forward plug (104b) of intermediate pre-hydrolysate solution out of reactor (Rl) into storage vessel (103) as illustrated in figure 1C.

In addition, suitably, a plug of displacement fluid (106b) is introduced into reactor (R5). The plug (106b) of displacement fluid being introduced in reactor (R5), suitably pushes forward plug (105a) that was residing in reactor (R5). Plug (105a), previously contained merely fresh second aqueous hydrochloric acid solution, but has in the meantime taken up saccharides from the solid material in reactor (R5) and has become an intermediate hydrolysate solution. Plug (105a) is pushed out of reactor (R5) into reactor (R4), thereby pushing forward plug (105b) of intermediate pre-hydrolysate solution out of reactor (R4) into reactor (R3). The plug (105b) of intermediate pre-hydrolysate solution, pushes forward plug (105c) that was residing in reactor (R3). Plug (105c), previously contained intermediate hydrolysate solution, but has now taken up sufficient saccharides and has become an aqueous second hydrolysate product solution. Such second hydrolysate product solution can also be referred to as a hydrolysate product solution. Plug (105c) of second hydrolysate product solution is pushed out from reactor (R3). Such second hydrolysate product solution can suitably be forwarded to one or more subsequent processes or devices, where optionally hydrochloric acid could be removed from the hydrolysate solution and recycled.

During this same second part of the cycle, residue (107) containing lignin can suitably be removed from reactor (R6) via solid outlet line (108) and reactor (R6) can be loaded with a new batch of dried wood chips (shown as (201) in figure 2A). The cycle has now been completed and all reactors have shifted one position in the reactor sequence. That is:

- reactor (R6) has now shifted into the position previously occupied by reactor (Rl);

- reactor (Rl) has now shifted into the position previously occupied by reactor (R2);

- reactor (R2) has now shifted into the position previously occupied by reactor (R3);

- reactor (R3) has now shifted into the position previously occupied by reactor (R4);

- reactor (R4) has now shifted into the position previously occupied by reactor (R5); and

- reactor (R5) has now shifted into the position previously occupied by reactor (R6).

As indicated, the above cycle takes about 8 hours. A subsequent cycle can now be started.

The situation wherein all reactors have shifted one position has been illustrated in Figure 2A. Figure 2A illustrates the start of a subsequent cycle, at a time "t+8 hours". The dried wood chips in what was previously reactor (R6) and is now reactor (Rl) can be flooded with a plug (204c) of intermediate pre hydrolysate solution withdrawn from the storage vessel (103). This is the same intermediate pre hydrolysate solution that was stored in such storage vessel (103) as plug (104b) of intermediate pre hydrolysate solution in the second part of the previous cycle, and illustrated in figure 1C. The subsequent cycle can be carried out in a similar manner as described above for the preceding cycle. Such is illustrated in figure 2B, where numerals (201), (202), (204a-d), (205a-c) and (206a-d) refer to features similar to the features referred to by numerals (101), (102), (104a-d), (105a-c) and (106a-d) in figure IB.

It is noted that all pre-hydrolysate and hydrolysate solutions in the above examples are suitably aqueous hydrolysate solutions, respectively aqueous pre-hydrolysate solutions.

Example 2

In this experiment, a sample of one batch woodchips (pinewood, Staatsbosbeheer, Netherlands) was subjected to a series of analysis and estimations.

The sample as mentioned above was subjected to an analysis on moisture and the amount of

hemicellulose and its composition and cellulose.

Next, a sample of the same batch of these woodchips were analysed on packed bed density, when loosely put in a reactor. Furthermore, a sample of the same batch of these woodchips were submitted to a particle size distribution.

Methods

The moisture was determined by placing the biomass sample in a laboratory oven at 105°C for 16 hours, and measuring the loss in weight.

The amount of various C5 and C6 sugars in the woodchips assumed to be representative of all hemicellulose present was determined by the method according to NREL LAP: NREL/TP-510-42618.

The method does not distinguish between glucose from hemicellulose and glucose from cellulose. All glucose from this determination was assumed to be of cellulose (in reality about 6% of glucose is part of hemicellulose, the rest is from cellulose.

The packed bed density was determined by packing 10 reactors of the same size and shape as are used for hydrolysis trials (i.e. 1.2 L volume, 530 mm tall 63 mm diameter), and measuring the weight of the amount of particles that can fit in.

The particle size distribution was determined by sieve analysis. For this, a sample of about 200 g was placed in a sieve stack with sieves having square holes of 100 mm, 63 mm, 45 mm, 31 mm, 25 mm, 16 mm, 8 mm and 3.15 mm. The sieve stack so loaded was shaked for 15 minutes. Below 3.15 are considered fines. Samples where then weighed back and the weight fraction between each size was recorded.

Results

The moisture content was 5.52%.

The result of the compositional analysis of the sample on dry basis was as in table 1. Remainder was unknown.

Table 1: compositional analysis, weight %

The packed bed density was 190 kg/m ³ .

The particle size distribution was as in table 2.

Table 2: particle size distribution

Example 3

Another sample of the same batch of woodchips as in example 2 was subjected to a hydrolysis trial process in a single flow-through reactor. The liquid flow-in was hydrochloric acid of 37%, and of the flow- out the amount of sugars in the liquid was measured, as percentage of the amount of hemicellulose.

Method and materials

In one tubular reactor (PVC, mounted vertically) with a size of 1.2 liter (dimensions as in example 2) 228g woodchips was packed. The reactor was equipped with sieve plates at the top and bottom. The inlet at the bottom was connected to a supply of fresh hydrochloric acid of 37%, the outlet at the top. Samples were taken from the flow out the top of the reactor at hour 0, 1, 2, 3, 4,5, 6, 7 and then after 24 hours of constant flow. The samples where then analysed for sugar content using an 1C. The hydrochloric acid was pumped into the reactor with a flow rate of 5 ml/min, for a total duration of 24 hours. Following the trial, the residual solid material was washed with water (20g water/g biomass) and then dried for 16h at 105°C. This was done to indicate the water holding capacity of the solid material, referred to as retained liquid, to serve as one of the input for calculating the conditons for main hydrolysis.

Results

The result was a curve of hemicellulose sugar recovery over time of the experiment, expressed as a percentage of the total amount of hemicellulose measured (as in example 2), starting at 0% and increasing to above 70%. The resulting curve is set out in figure 3.

This gives an indication on how the amount of hydrochloric acid required theoretically relates to such on the woodchips in an actual hydrolysis experiment, and differences are e.g. due to particle size, particle porosity, and others. Following this testing, the resulting solid mass was also shown to have 3 g residual HCI/g solid material (retained liquid).

Example 4

Of the sample analysed in example 2 and experimented on in example 3 an estimation of the theoretical amount of hydrochloric acid (37% for hemicellulose, 42% for cellulose) required for full hydrolysis and separation of hemicellulose and cellulose sugars was made. This was done by taking the result from the curve in figure 3 of example 4 and the retained liquid mass an assessing the amount of HCI 37% required to hydrolyse and wash the majority of hemicellulose sugars out sugars to achieve a 90% (g glucose/g total sugar) solution in main hydrolysis. A calculation was then made to estimate the amount of 42% HCI required to keep the acid concentration in main hydrolysis high enough (for safety >41% HCI) to complete the cellulose hydrolysis.

Results

Reading from the curve in figure 3, for the given sample a total of 22g HCI 37%/g biomass was required for the desired degree of hydrolysis. Using for the retained liquid an amount of 3 g/g it was then calculated that a minimum of 12 g HCI 42/g biomass was required to bring the average concentration in main hydrolysis to above 41% HCI, which is a safe level for ensuring hydrolysis goes to completion. g HCI

3 * 37% HCl + x * 42% HCl

g biomass

41% HCI =

3 + x Given a reactor volume of 1.2 L, a biomass fill of 220 g, this means that the retained liquid mass (approx. 660g) can be used to estimate the displacement fluid volume which is flowed to the next reactors in case a set up as in example 1 is used for production. This was calculated to be equivalent to 2.75 g HCI 37%/g biomass pushed to the next reactor in step 106a as in example 1. A rule-of thumb developed from experience with several wood chip types suggests that 5 g HCI 42/g biomass was left in the retained liquid fraction following main hydrolysis allowing a similar calculation to be made here.

This was converted to a flow scheme as presented in example 1 by using a fresh plug of 5.5 g HCI 37%/g biomass in 104a, with 3 reactors in sequence in pre hydrolysis and 3 reactors in sequence in main hydrolysis and a plug of 3g HCI 42%/g biomass in 105a. Each reactor in pre-hydrolysis therefore sees flow from 3 fresh plugs and 2 displacement fluid plugs, which makes up the 22 g HCI/g Biomass as suggested. The experiment was run over 10 fresh biomass fills, and produced the sugar concentation profiles outlined in figure 4, which achieved between 88-95% glucose purity in each main hydrolysate product. Fluctuations are due to experimental errors.