The present invention relates to a preparation of carrot cell wall material and a method for preparing a preparation of carrot cell wall material. It also relates to a composition comprising an aqueous phase and a preparation of carrot cell wall material and a method to prepare such a composition. The invention also provides use of a preparation of carrot cell wall material to modify the rheological properties of an aqueous phase and to use of such a preparation to structure an aqueous phase.

BACKGROUND TO THE INVENTION

Control over the rheological properties of compositions comprising an aqueous phase is desirable in many applications. This particularly holds true for food compositions. Many foodstuffs have or require a structured aqueous phase. Some foodstuffs have thickened or well-suspending water phases by their very nature, whereas in others consumer appreciation or product stability (for example against sedimentation or coalescence) depend on additional means to control their rheology.

Denatured, gelatinised starch is a well-known structurant for aqueous systems.

However, it directly affects the caloric value of any composition comprising it.

Moreover, many modified starches are not perceived as very natural by many consumers.

Citrus fibres are also known as structuring agents for aqueous phases. However, these often require extensive processing before they are suitable for use as structuring agents. Such processing is relatively costly and may reduce the perceived naturalness of a composition comprising such fibres. Other cellulose-based, plant-derived structuring agents have also been developed. However, such materials often exhibit related issues. For example, WO 2014/142651 discloses the use of particulate cellulose material derived from vegetable pulp - in particular from sugar beet pulp - for many different types of rheological modification, but the material requires chemical and or enzymatic treatment to degrade and or extract pectin and hemicellulose. Likewise, WO 2014/147393 discloses a process for preparing cellulose containing particles from plant material using a peroxide agent. The bleaching step is believed to partially degrade pectins and hemicelluloses.

In view of the above it is an object of the present invention to provide plant-derived structuring agents that do not depend on bleaching, enzymatic treatment or like processes to provide desirable and tuneable rheological characteristics.

It is a particular object of the present invention to provide a natural material capable of modifying the rheology of aqueous phases, structuring aqueous phases, or modifying their texture, in particular to provide such a natural material of low caloric value.

It is another object of the present invention to provide such a material based on plant cell wall material. It is a more particular object to provide such a material without the need for bleaching steps or other chemical or enzymatic treatments in its manufacture.

Therefore it also is an object of the present invention to provide a method for preparing such a material, in particular a method that does not rely on bleaching steps or other chemical or enzymatic treatments in order to provide a plant-based material with excellent rheology-modifying and/or structuring properties.

It is another object of the present invention to provide compositions, in particular food compositions comprising an aqueous phase wherein the aqueous phase has controllable rheological properties, is structurable, and wherein the structurant is plant- based.

DEFINITION OF THE INVENTION

We have surprisingly found that one or more of these objects can be achieved by the present invention. In particular, we found that carrot cell wall material with excellent structuring and/or rheology-modifying characteristics can be prepared from carrots, provided that solutes and other non-cell wall components are removed from the cell wall material to a sufficient degree. Such a preparation of carrot cell wall material can be prepared using a relatively simple process that does not depend on enzymatic treatment or bleaching. Thus, a carrot cell wall material that is characterised in terms of its particle size and in terms of the ratios of several saccharide monomers of the fibrous material making up the cell wall material was found to be highly suitable to structure aqueous phases, tailor their rheological properties and/or control their suspending properties.

Therefore, according to a first aspect, the invention provides a preparation of carrot cell wall material, wherein:

• the cell wall material is in particulate form and comprises particles having a size of between 25 and 500 μηη;

· the cell wall material has a ratio of galacturonic acid monomers to bound glucose monomers of at most 1 .0; and

• the cell wall material has a ratio of galactose monomers to bound glucose

monomers of at least 0.15;

• wherein the monomer ratios are by weight.

A preparation of carrot cell wall material displaying the above desirable characteristics can be made by a relatively simple process directed at removing solutes and other non-cell-wall components from the cell wall material by a method involving heating and pureeing followed by a washing step. Therefore, according to a second aspect, the invention provides a method for preparing a preparation of carrot cell wall material according to the invention, wherein the method includes the steps of

a. providing carrot, optionally in sliced, diced or shredded form;

b. subjecting the carrot to a heat treatment at a temperature of between 85°C and 1 10°C for a period of at least 20 minutes;

c. pureeing the carrot material before or after said heat treatment;

d. washing the heated and pureed carrot material by contacting it with water and separating the carrot material from at least part of the wash water by subjecting it to a liquid/solid separation treatment; and

e. optionally repeating the washing step.

According to a third aspect, the present invention provides a preparation of carrot cell wall material obtainable by the method according to the invention. The preparation of carrot cell material of the present invention is highly suitable to modify the rheological properties of aqueous phases. Therefore, according to a fourth aspect, the present invention provides a composition comprising an aqueous phase and the preparation of carrot cell wall material according to the invention, wherein the preparation is dispersed in said aqueous phase.

Likewise, according to a fifth aspect, the present invention provides a method for preparing a composition comprising a structured aqueous phase, including the step of dispersing the preparation of carrot cell wall material according to the invention into an aqueous medium so as to form said structured aqueous phase.

According to a sixth aspect, the present invention provides use of a preparation of carrot cell wall material according to the invention in a composition comprising an aqueous phase to modify the rheological properties of said aqueous phase.

According to a seventh aspect, the present invention provides use of a preparation of carrot cell wall material according to the invention in a composition comprising an aqueous phase to structure said aqueous phase. BRIEF DESCRIPTION OF FIGURES

FIGURE 1 provides a CSLM micrograph of the carrot cell wall material of Example 18. FIGURE 2 provides a CSLM micrograph of the carrot cell wall material of Example 19. FIGURE 3 provides a CSLM micrograph of the carrot cell wall material of Example 21. DETAILED DESCRIPTION OF THE INVENTION

Any feature of one aspect of the present invention may be utilised in any other aspect of the invention. The word "comprising" is intended to mean "including" but not necessarily "consisting of or "composed of." In other words, the listed steps or options need not be exhaustive. It is noted that the examples given in the description below are intended to clarify the invention and are not intended to limit the invention to those examples per se. Similarly, all percentages are weight/weight percentages unless otherwise indicated. Moreover, weight percentage (wt. %) is based on the total weight of the product unless otherwise stated. Except in the operating and comparative examples, or where otherwise explicitly indicated, all numbers in this description indicating amounts of material or conditions of reaction, physical properties of materials and/or use are to be understood as modified by the word "about". Unless specified otherwise, numerical ranges expressed in the format "from x to y" are understood to include x and y. When for a specific feature multiple preferred ranges are described in the format "from x to y", it is understood that all ranges combining the different endpoints are also contemplated. For the purpose of the invention ambient temperature is defined as a temperature of about 20 degrees Celsius. Preparation of carrot cell wall material

The present invention relates to a preparation of carrot cell wall material.

In the context of the present application, carrot cell wall material is understood as material derived from the cell walls of the tissue of carrot (i.e. the taproot of Daucus carota). In general, such cell wall material, as understood by the person skilled in the art, is predominantly made up of cellulose, hemicellulose and pectin.

The the cell wall material is in particulate form and comprises particles having a size of between 25 and 500 μηη. For the purpose of the present invention, the particle size is determined by wet sieving. Thus a particle size below 500 μηη means that the particles pass a sieve with an aperture size of 500 μηη. Likewise, a particle size above 25 μηη means that a particle does not pass a sieve with a pore size of 25 μηη. In wet sieving, an aqueous dispersion of the cell wall material at a suitable concentration is passed through a sieve, as detailed hereinbelow. Without wishing to be bound by theory, it is believed that cell wall material within the size ranges is capable of acting as a rheology modifier by virtue of its ability to build extended networks throughout an aqueous phase. Fragments that are too large do not significantly contribute to such network formation. Moreover they may impart other undesirable characteristics, as they may lead to a grainy, gritty structure. Especially for applications in food products, these would be undesirable.

In view of the above, the preparation of carrot cell wall material preferably does not contain large amounts of particles having a size outside the specified range. Therefore, preferably at least 80 wt-%, more preferably at least 90 wt-% and even more preferably at least 95 wt-% and still more preferably more preferably at least 98 wt-% by dry weight of the carrot cell wall material in the preparation passes a sieve of 500 μηη pore size.

Preferably, the cell wall material comprises particles having a size not more than 400 μηη, more preferably not more than 300 μηη, even more preferably not more than 200μηι and still more preferably not more than 150 μηη. Thus, it is preferred that at least 80 wt-% by dry weight of the carrot cell wall material in the preparation passes a sieve of 400 μηη pore size, more preferably of 300 μηι, even more preferably of 200 μηη and still more preferably of 150 μηη pore size.

Cell wall fragments that are below the lower size limit are believed to be too small to provide the desired benefits as they cannot contribute to the network formation. In this respect it is hypothesized that particles including one or more intact cell wall enclosures and/or one or more fragments of such enclosures that still retain their native spheroidal or cotyloid shape are particularly effective in imparting the desired rheological modifications. Therefore particles comprising such intact enclosures and/or cotyloid fragments are preferred. In general, typical carrot cells have a diameter of about 40 micrometres. Thus, for example fragments of cell wall material with a diameter of about 25 μηη are believed to sufficiently retain the cotyloid shape related to the excellent structuring properties. The advantageous curvature associated with intact cell wall enclosures or cotyloid fragments is believed to be non-discernible when cell wall fragments are much smaller than about 25 μηη in size. Such small fragments typically become more platelet-like.

Therefore, preferably no more than 30 wt-%, more preferably no more than 25 wt-%, even more preferably no more than 20 wt-%, still more preferably no more than 15 wt- %, and yet more preferably no more than 10 wt-% by dry weight of the carrot cell wall material in the preparation passes a sieve of 25 μηη pore size.

It is even more preferred that no more than 30 wt-%, more preferably no more than 25 wt-%, even more preferably no more than 20 wt-%, still more preferably no more than 15 wt-%, and yet more preferably no more than 10 wt-% by dry weight of the carrot cell wall material in the preparation passes a sieve of 40 μηη pore size.

Suitable sieves include for example stainless steel woven wire laboratory test sieves of the appropriate aperture size available from Endecotts Ltd (London, England) as described in the Methods hereinbelow. Separation of particles of carrot cell wall material with sizes above and below 25 μηη, respectively, may also suitably be carried out using a filter cloth with the appropriate pore size, for example a Miracloth filter (Calbiochem) as described in the Methods hereinbelow.

Preferably at least 50 wt-% of the particles comprise one or more intact cell wall enclosure, more preferably at least 75 wt-% and even more preferably at least 90 wt-%, wherein the percentage is by dry weight of the particles. In the context of the present invention, intact cell wall enclosures are understood to be enclosures that retain the spheroidal shape of the original carrot cells, though the solutes and other cell-constituents have been removed at least in part. The presence of intact cell wall enclosures in the preparation of carrot cell wall material can easily be observed using microscopic techniques that are suitable for visualising cell walls as known to the skilled person. Confocal scanning microscopy is an example of such a technique as explained hereinbelow.

The particle size of the cell wall material can for instance be controlled by conventional methods known to the skilled person, including well-established comminution methods and size selection methods.

Galacturonic acid to glucose monomer ratio

It was surprisingly found that the advantageous rheological or structuring properties of the preparation of cell wall material of the present invention is related to the ratio of the galacturonic acid and the glucose monomers present in the preparation.

Thus, the carrot cell wall material in the preparation has a ratio of galacturonic acid monomers to bound glucose monomers of at most 1 .0, wherein the monomer ratio is by weight of the monomers. The ratio of galacturonic acid monomers to bound glucose monomers is preferably at most 0.98, more preferably at most 0.95, and even more preferably at most 0.91 by weight of the monomers. The monomer ratio can be determined by a method based on NMR spectroscopy and Saeman hydrolysis as explained in the Methods section hereinbelow.

In the context of the present invention, "bound glucose monomers" refer to those glucose monomers that are part of the polysaccharides in the cell wall material. Such bound glucose is predominantly present in the cellulose and hemicellulose. For a given portion of carrot material, the amount of bound glucose monomers may be lower than the total amount of glucose monomers, because by nature, carrots typically also comprise mono- and oligosaccharides, such as glucose and sucrose. The ratio of galacturonic acid to bound glucose of the cell wall material in the present preparation is lower than that in native, untreated carrot cell wall material. The desired ratio can easily be arrived at by the method according to the second aspect of the present invention. It is believed that galacturonic acid monomers are predominantly present in the pectin contained in the cell wall material. In contrast, the bound glucose is predominantly present in the cellulose and hemicellulose. Without wishing to be bound by theory, it is believed that owing to its greater water-solubility, the ratio of pectin with regard cellulose and hemicellulose decreases during manufacture of the preparation of carrot cell wall material of the present invention. Other solutes

It is believed that the preparation of carrot cell wall material owes its surprising advantageous properties to the removal of solutes to a sufficient degree. The solute content of the carrot cell wall material can suitably be quantified in terms of the conductivity of a dispersion of the cell wall material, since such conductivity typically depends on the level of water-soluble electrolytes in the preparation of carrot cell wall material. Therefore, it is preferred that upon preparation of an aqueous dispersion comprising the carrot cell wall material in an amount corresponding to a dry matter content of 1 .00 wt-% with regard to the total dispersion, said dispersion has a conductivity at 20 °C of at most 300 μδ/οηη (microsiemens per centimeter), more preferably at most 200 μ8/θΓη and even more preferably at most 100 μδ/οηη, yet more preferably at most 50 μ8/οη"ΐ, still more preferably at most 20 μδ/οηη and even still more preferably at most 15 μδ/cm, wherein the dispersion essentially consists of the preparation of carrot cell wall material and demineralised water.

Likewise, the solute content of the carrot cell wall material can also suitably be quantified in terms of its Brix value. Therefore, it is preferred that upon preparation of an aqueous dispersion comprising the carrot cell wall material in an amount

corresponding to a dry matter content of 5.00 wt-% with regard to the total dispersion, said dispersion has a Brix value of 2 °Brix or less, more preferably 1 °Brix or less, even more preferably 0.2 °Brix or less, and still more preferably 0.1 °Brix or less, wherein the Brix value is determined by the refractive index-based method as known to the skilled person.

Other monosaccharide ratios

It was surprisingly found that the advantageous rheology-modifying and structuring properties of the preparation of carrot cell wall material can surprisingly be obtained under relatively mild conditions, without the need of an intensive bleaching step. It is believed that typical bleach treatments, such as those in WO 2014/147393, lead to substantial degradation of hemicellulose. Consequently, the ratio of galactose monomers to bound glucose monomers is very low in such bleached material, since galactose monomers are predominantly present in hemicellulose. In contrast to such bleached materials, the cell wall material of the present preparation of carrot cell wall material has a ratio of galactose monomers to bound glucose monomers of at least 0.15, wherein the monomer ratio is by weight of the monomers. The ratio of galactose to bound glucose is preferably at least 0.17, more preferably at least 0.20, and even more preferably at least 0.25 by weight of the monomers. Likewise, the preparation of carrot cell wall material of the present invention preferably has a ratio of galacturonic acid monomers to bound glucose monomers of at least 0.45, more preferably at least 0.50, and even more preferably at least 0.55, by weight of the monomers. Consequently, the carrot cell wall material preferably has a ratio of galacturonic acid monomers to bound glucose monomers of between 0.98 and 0.45, more preferably between 0.95 and 0.50, and even more preferably between 0.91 and 0.55 by weight of the monomers. Similarly, the preparation of carrot cell wall material of the present invention preferably has a ratio of arabinose monomers to bound glucose monomers of at least 0.10, more preferably at least 0.12, even more preferably at least 0.14 and still more preferably at least 0.16 by weight of the monomers. These monomer ratios can be determined by the same method as the ratio of galacturonic acid to bound glucose.

Beta-carotene

Beta-carotene is the compound that gives most Daucus carota cultivars their characteristic orange colour. It is believed that during typical methods for making the preparation of carrot cell wall material of the present invention, the solute removal also affects the level of beta-carotene in the preparation. Therefore, the preparation of cell wall material preferably comprises beta-carotene in an amount of at most 0.04 wt-%, more preferably at most 0.03 wt-%, even more preferably at most 0.02 wt-%, still more preferably at most 0.01 wt-%, even still more preferably at most 0.005 wt-% and yet more preferably at most 0.004 wt-% with respect to the dry weight of the preparation. Conversely, severe bleaching steps - which are not required for the present preparation - would typically remove substantially all beta-carotene. Therefore the preparation of cell wall matter preferably comprises beta-carotene in an amount of at least 0.0001 wt-%, more preferably at least 0.001 wt-% with respect to the dry weight of the preparation. Thus, the preparation preferably comprises between 0.04 wt-% and 0.0001 wt-%, more preferably between 0.03 wt-% and 0.001 wt-% of beta-carotene with respect to the dry weight of the preparation. Preparation of carrot cell wall material in dry form

In one preferred embodiment, the preparation of carrot cell wall material according to the invention is in dry form. The preparation in dry form need not be free of water, but has a dry appearance. Consequently, the preparation of carrot cell wall material preferably comprises 15 wt-% or less, more preferably 12 wt-% or less, and even more preferably 10 wt-% or less of water, with regard to the total weight of the preparation. Thus, the dry matter content of the preparation in dry form is preferably at least 85 wt- %, more preferably at least 90 wt-% by weight of the total preparation. Conversely, it is preferred that the preparation comprises the carrot cell wall material in an amount of at least 50 wt-%, more preferably at least 75 wt-%, even more preferably at least 90 wt-%, and still more preferably at least 95 wt-%, wherein the weight percentage is the percentage of the dry weight of the carrot cell wall material with respect to the total dry weight of the preparation.

The preparation in dry form is particularly advantageous from a perspective of storage stability and weight efficiency during transport and storage. The advantageous properties of the preparation relate to its behaviour in dispersions, typically aqueous dispersions. Therefore, the preparation in dry form is preferably such that it easily regains its rheology modifying or structuring properties upon redispersing in a liquid medium, in particular water. The preparation in dry form is typically obtainable by a process involving a drying step. Such a drying step is preferably carried out in such a way that the beneficial properties of the preparation are retained. It is believed that the structuring and or rheology-modifying properties are optimally retained if the drying is carried out whilst collapse of the intact cell wall enclosures and cotyloid fragments is minimised or even avoided. Therefore, it is preferred that in the preparation of carrot cell wall material of the present invention in dry form, the outer volume of the intact cell wall enclosures is retained with regard to the volume of the enclosures before the drying step. It is even more preferred that the particles of carrot cell wall material are redispersible in water by the application of only mild shear, for example by stirring or by agitation using equipment like a Silverson blender. Suitable drying steps to obtain the preparation of the present invention in dry form include techniques like spray drying, freeze-drying, or drying methods based on solvent-exchange. Preferably the preparation of carrot cell wall material in dry form comprises the carrot cell wall material in an amount of at least 50 wt-%, more preferably at least 75 wt-%, even more preferably at least 90 wt-%, and still more preferably at least 95 wt-%, wherein the weight percentage is the percentage of the dry weight of the carrot cell wall material with respect to the total dry weight of the preparation. It is even more preferred that the dry matter of the preparation of carrot cell wall material in dry form essentially consists of carrot cell wall material. The preparation in dry from may also comprise an additive directed at retaining the non-collapsed structure of the cell wall material during drying or at facilitating its redispersion upon use. Suitable additives are known to the skilled person. It is preferred that the additive is suitable for use in food applications. In view of its applicability, it is preferred for the preparation in dry form, that re- dispersion of the preparation of carrot cell wall material in water using the Standard Silverson redispersion treatment defined hereinbelow so as to form an aqueous dispersion comprising the preparation of carrot cell wall material in an amount of 1.00 percent by weight of dry material with respect to the total weight of the dispersion yields a shear storage modulus G' of at least 50 Pa, more preferably at least 100 Pa even more preferably at least 200 Pa, and still more preferably at least 500 Pa for said aqueous dispersion, wherein G' is measured as G'(5 min eq.) as defined herein.

Likewise, it is preferred for the preparation in dry form, that re-dispersion of the preparation of carrot cell wall material in water using the Standard Silverson

redispersion treatment defined hereinbelow followed by dilution and manual mixing so as to form an aqueous dispersion comprising the preparation of carrot cell wall material in an amount of 0.3 percent by weight of dry material with respect to the total weight of the dispersion yields a self-suspending capacity after 24 hours of at least 25%, more preferably at least 50%, even more preferably at least 70%, and still more preferably at least 80% for said aqueous dispersion.

Preparation of carrot cell wall material in wet form

According to another preferred embodiment, the preparation of carrot cell wall material according to the invention is in wet form. In wet form, the preparation typically comprises one or more solvents, including water. The dry matter content of the preparation in wet form may vary. Thus the preparation may e.g. be in the form of a relatively dilute aqueous dispersion of carrot cell wall material, but it may also be in the form of a concentrated slurry (such as may be formed when the cell wall matter is washed and filtered) or for instance a relatively dense pellet formed during

centrifugation of more dilute dispersions. Even such relatively dense forms of the preparation rarely contain more than 25 wt-% of carrot cell wall material (in terms of dry matter) with regard to the total preparation. Consequently, the preparation of carrot cell wall material when in wet form, preferably comprises between 0.1 wt-% and 25 wt-%, more preferably between 0.25 wt-% and 20 wt-%, even more preferably between 0.5 wt-% and 15 wt-%and still more preferably between 1 wt-% and 10 wt-% of carrot cell wall material in terms of dry weight with respect to the total weight of the preparation.

It is preferred that the preparation of carrot cell wall material of the present invention is in the form of an aqueous dispersion of the cell wall particles.

In some applications (for example if the preparation of carrot cell wall material is in the form of an aqueous dispersion), it preferably comprises between 0.1 wt-% and 15 wt- %, more preferably between 0.25 wt-% and 13 wt-%, even more preferably between 0.5 wt-% and 10 wt-% and still more preferably between 1 wt-% and 5 wt-% by dry weight of the carrot cell wall material with respect to the total weight of the preparation. In other applications (for example if the preparation of carrot cell wall material in wet form is in the form of a concentrated slurry, a filtration residue or a centrifuge pellet) it preferably comprises between 5 wt-% and 25 wt-%, more preferably between 10 wt-% and 23 wt-%, even more preferably between 15 wt-% and 20 wt-% by dry weight of the cell wall material with respect to the total weight of the preparation.

The preparation in wet form may suitably be applied to structure or rheologically modify aqueous phases by dispersing the preparation therein. Therefore, it is preferred for the preparation in wet form, that dispersion of the preparation of carrot cell wall material in water using the Standard Silverson redispersion treatment defined hereinbelow so as to form an aqueous dispersion comprising the preparation of carrot cell wall material in an amount of 1 .00 percent by weight of dry material with respect to the total weight of the dispersion yields a shear storage modulus G' of at least 50 Pa, more preferably at least 100 Pa, even more preferably at least 200 Pa and still more preferably at least 500 Pa for said aqueous dispersion, wherein G' is measured as G'(5 min eq.) as defined herein.

Likewise, it is preferred for the preparation in wet form, that dispersion of the

preparation of carrot cell wall material in water using the Standard Silverson

redispersion treatment defined hereinbelow followed by dilution and manual mixing so as to form an aqueous dispersion comprising the preparation of carrot cell wall material in an amount of 0.3 percent by weight of dry material with respect to the total weight of the dispersion yields a self-suspending capacity after 24 hours of at least 25 %, more preferably at least 50 %, even more preferably at least 70%, and still more preferably at least 80% for said aqueous dispersion.

Method for preparing the preparation of carrot cell wall material

According to the second aspect, the present invention relates to a method for preparing a preparation of carrot cell wall material according to the first aspect of the invention, comprising the steps detailed hereinabove.

The preferences for the preparation of carrot cell wall material according to the first aspect of the invention also apply to the method according to the second aspect of the invention.

During step a. of the method, the carrot may be provided in any form that is suitable for subjecting to the heat treatment of step b. Therefore, the carrot is preferably provided in sliced or diced or shredded form. The carrot starting material may be subjected to such a slicing/dicing or shredding treatment directed specifically at making it suitable for the present application, but may also be the result of earlier treatments. The latter may for instance be the case if the carrot is sourced from a waste stream such as peelings, scratchings, or press-cakes. From a sustainability point of view it is preferred that the carrot is sourced from a waste stream of another process, in particular another process directed at obtaining a food product or beverage from carrot.

Step b of the method includes a heat treatment at a temperature of between 85°C and 1 10°C for a period of at least 20 minutes. Preferably, the temperature is between 90 °C and 105°C, more preferably between 92°C and 100° and even more preferably between 93°C and 98°C. The heat treatment is preferably for a period of between 20 and 45 minutes. In order for the washing step to be efficient at removing water-soluble solutes and other non-cell wall components, the carrot material should be subjected to a pureeing step c before said washing step. The carrot material may be subjected to the pureeing step before or after the heat treatment. In some cases, the depending on the source of the carrot material, said pureeing may already have taken place in the step of providing the carrot. Pureeing may be carried out by any suitable pureeing equipment as known to the skilled person, including for example Kenwood or Waring or Thermomix blenders, or Silverson mixers.

The washing step d is directed at removing solutes and other non-cell wall components from the carrot cell wall material to a sufficient degree. The liquid/solid separation treatment may be any suitable separation treatment known to the skilled person.

Examples of suitable separation treatments include centrifugation, decantation, filtration, belt-pressing. The washing treatment may be a treatment that is suitable for batch-wise or for continuous operation. Depending on the type of washing, a single step of contacting with water and subsequent separation may suffice, whilst in other types, the washing needs to be repeated in order to obtain a preparation of carrot cell wall material according to the present invention.

Therefore, the washing step is preferably carried out, optionally repeatedly, until the wash water that is separated from the carrot cell wall material has a Brix level of not more than 2 °Brix and/or a conductivity at 20°C of not more than 300 μ8/οη"ΐ.

Here, it is particularly preferred that said Brix level is not more than 1 °Brix, more preferably not more than 0.2 °Brix and even more preferably not more than 0.1 °Brix, wherein the Brix is determined by the refractive index method as known to the skilled person. The conductivity of the wash water that is separated from the carrot cell wall material not only depends on the level of solutes present in the carrot material, but also on the background conductivity of the water before it is contacted with the carrot material. Typical process water has a conductivity at 20 °C of between 300 με/οηη and 50 μ8/οπ"ΐ . Therefore, it is particularly preferred that if the water that is contacted with the carrot material has a conductivity before said contacting of between 300 μ8/θΓη and 50 μ8/οη"ΐ , the washing step is carried out, optionally repeatedly, until the conductivity of the wash water after separation has a conductivity at 20°C that is no more than 150%, more preferably no more than 120%, even more preferably no more than 1 10% and still more preferably no more than 105% of the conductivity of the water before contacting it with the carrot material. It is also preferred that the washing step is carried out, optionally repeatedly, until the conductivity of the wash water after separation has a conductivity at 20°C of at most 200 μδ/cm, more preferably at most 100 μ8/οη"ΐ, yet more preferably at most 50 μδ/cm, still more preferably at most 20 μ8/θΓη and even still more preferably at most 15 μδ/cm.

The method for preparing the preparation of carrot cell wall material may suitably include more steps between the initial heating step and the final washing step. Such additional steps may include subjection to one or more comminution treatments, one or more additional heat treatments and one or more additional washing treatments, for instance in order to arrive at a preparation with the appropriate particle size and with the appropriate ratio of galacturonic acid monomers to bound glucose monomers and the approriate ratio of galactose monomers to bound glucose monomers. Thus, the method preferably includes one or more additional steps in which the carrot cell wall material is subjected to a comminution treatment and wherein these comminution treatments take place between said heating step and the last washing step. The additional steps may be present in any suitable order and number. Suitable

comminution steps may for example include relatively low shear blending, pureeing, milling or grinding treatments, or high shear treatments, including for instance high pressure homogenisation.

The method for preparing the preparation of carrot cell wall material may also include additional steps after the final washing step. Thus, the method may optionally include a comminution step after the final washing step, for example in order to reduce the size of the particles to the required size.

If the preparation of carrot cell wall material is in dry form, the method typically also includes a drying step. Such a drying step is preferably carried out in such a way that the beneficial properties of the preparation are retained. It is believed that the structuring and or rheology-modifying properties are optimally retained if the drying is carried out whilst collapse of the intact cell wall enclosures and cotyloid fragments is minimised or even avoided. Suitable drying steps to obtain the preparation of the present invention in dry form include techniques like spray drying or freeze-drying. Alternatively, such a drying step may include a solvent-exchange process, including the steps of

a) contacting the wet carrot cell wall material with a non-aqueous water-miscible solvent

b) subjecting the resulting dispersion to a liquid/solid separation treatment c) optionally repeating the contacting and separation treatments

d) drying the separated carrot cell wall material.

The non-aqueous water-miscible solvent preferably is a solvent in which the cell wall material is poorly dispersible. For example, it is preferred that the solvent is an alcohol, more preferably it is ethanol or isopropanol or a mixture thereof. Advantageously, a solvent mixture that is sufficiently enriched in non-aqueous solvent (and

correspondingly low in water) may be evaporated from the cell wall material whilst avoiding collapse of its structure, without the need for spray-drying or freeze-drying.

The preparation of carrot cell wall material according to the invention is preferably obtainable by the method according to the second aspect of the invention.

Compositions comprising the preparation

According to the fourth aspect, the invention provides a composition comprising aqueous phase and the preparation of carrot cell wall material according to the invention, wherein the preparation is dispersed in said aqueous phase. The composition may be in any suitable format. The preparation of carrot cell wall material is particularly suitable for use in edible compositions. Therefore, the composition preferably is a food composition. It is preferred that the composition, in particular when it is a food composition, is a composition of liquid, gelled, paste-like, spoonable, or semi-solid consistency. Thus, the food composition may for example be a soup, sauce, dressing, condiment, beverage, paste, ice cream, desert, or a dairy-based product. The preparation of carrot cell wall material is highly suitable to modify the rheological properties of the aqueous phase. Thus, it may provide it for example with the desired viscosity, suspending capacity, thickness, shear storage modulus, shear loss modulus, yield stress, Stevens value, or Bostwick value. Likewise, the preparation of the invention is highly suitable as a texture-modifier, which can impart an aqueous phase or entire product with the desired textural properties, including for instance

smoothness, pulpiness, or graininess. It is also highly suitable to structure the aqueous phase. Without wishing to be bound by theory, it is believed that such structuring is due to the ability of the particles of carrot cell wall material to form a network of throughout an aqueous phase. Therefore, the aqueous phase in the composition preferably is a structured aqueous phase.

The composition preferably comprises the preparation of carrot cell wall material in an amount of between 0.1 wt% and 10 wt%, more preferably between 0.2 and 5 wt%, even more preferably between 0.3 and 4 wt% and still more preferably between 0.4 and 3 wt% with regard to the weight of the aqueous phase.

In view of the above, the present invention also relates to a method for preparing a composition comprising a structured aqueous phase, including the step of dispersing the preparation of carrot cell wall material according to the invention into an aqueous medium so as to form said structured aqueous phase.

Any feature that is preferred in the preparation according to the first aspect, the method for preparing the preparation according to the second aspect, or the composition according to the third aspect of the invention is likewise preferred in this method according to the fourth aspect of the invention.

Use of the preparation of carrot cell wall material

Any preference described hereinabove with regard to the preparation according the invention applies equally to the use of the preparation of carrot cell wall material according to the sixth and seventh aspect.

In particular, the use according to the sixth aspect preferably is use of a preparation of carrot cell wall material according to the invention in a composition comprising an aqueous phase to provide said aqueous phase with a shear storage modulus of at least 50 Pa more preferably at least 100 Pa, even more preferably at least 200 Pa and still more preferably at least 500 Pa, wherein the shear storage modulus is measured as G'(5 min eq) as described herein.

METHODS

Rheoloav measurements

Samples were stored 24h at 4°C prior to the analysis of their rheological properties. A standard protocol was used to measure the shear storage modulus G' of the carrot cell wall material (cwm) samples. A stress-controlled rheometer (TA-instruments AR- 2000ex, unless indicated otherwise) with plate-plate geometry has been used. The top plate was a 40mm fixed plate sand blasted and a gap of Ι ΟΟΟμηη. First, the samples are gently stirred manually before being put on the plate. Then the standard protocol is started which consists of three steps.

· The equilibration step: the sample is subjected to a time sweep test at very low oscillatory shear for 5 min, the so-called equilibration time, at 1 Hz and a 0.1 % strain. The G' value after 5 min equilibration is taken as value called G' (5 min eq).

• Continuous ramp: shear rate varies from 0.1 s ^"1 to 500s ^"1 for 2min in logarithmic steps (the viscosity measured in this step is given at 0.1 at 1 and 10/s) and then back from 500 s ^"1 to 0.1 s ^"1 again for 2 minutes (viscosity measured in this step is presented using the Herschel-Bulkley equation as indicated in the Examples).

• Strain sweep: the amplitude of the deformation is represented by % strain and it varies from 0.1 % to 200% in logarithmic steps, with a fixed frequency of 1 Hz. The maximum value of the G' that is measured during the strain sweep measurement is taken as G' max at strain sweep (G' str. sw.).

All the steps are performed at 20 °C. For each experiment, G' of all samples was measured at least 2 times. In case of a discrepancy, a third measurement was done.

Self suspending capacity

To measure the % self suspending capacity of the fibers, dilutions of fiber samples in demineralised water were made at 0.3% DM yielding a total dispersion 100g. The dispersion was placed in a 100ml measuring cylinder. Samples were mixed manually and stored for 1 h and for 24h in order to observe the % volume occupied by the suspended fibers after 1 h and 24h, respectively. Particle size/filtration methods

Miracloth filter (from Calbiochem, supplied by Millipore, Typical pore size: 22-25 μηη, composed of rayon-polyester with an acrylic binder) was used for washing carrot cell wall material in water and obtaining a dispersion with wherein the particles of carrot cell wall material are larger than 25μη"ΐ. This was done by first adding 500g of a processed slurry of carrot cell wall material (between 1 -4% w/w) on the Miracloth filter using a Buchner funnel and adding portions of about 500ml demineralised water on top of the slurry and gently mixing the carrot cwm and water using a spoon. The filtrate was eluted by gravity. The washing of the cell wall material was done until the filtrate became almost colourless (about 5L of demineralised water was used for washing). The filtrate was removed and the residue was used further. As a result of washing the carrot cwm particles in the residue after Miracloth filtration are larger than 0.025 mm. Secondly, the particle size of the cell wall material was measured by filtration of a dispersion of carrot cell wall material in water through a sieve with 0.500 mm aperture size (Stainless steel woven wire laboratory test sieve from Endecotts Ltd, London, England). Filtration of a dispersion of carrot cwm (e.g. 1 .5 % DM) was performed by adding portions of about 500ml demineralised water on top of the slurry and gently mixing the carrot cwm and water using a spoon. The filtrate was eluted by gravity. The % residue was analysed based on % dry weight of the residue divided by the total dry weight.

Conductivity measurement

For the conductivity measurements, the samples containing the dispersed cwm were centrifuged for 30min at 6000rpm. A volume of 10 ml of the supernatant was then used to measure the conductivity using a standard bench top conductivity meter (Schott, model CS 855).

Refractive index and Brix value

The Brix value of a sample was measured using a digital Pocket refractometer PAL1 -RI (Atago, Tokyo, Japan). By applying about 0.3 mL of sample onto the measuring surface, the instrument displays the refractive index, the Brix and the temperature of the sample. Dry matter

For dry matter (DM) analysis, a sample of approximately 20g was taken and exposed overnight to a temperature of 80°C in an oven (Memmert) and then the dried sample was weighed to provide the total solid content (g). The % w/w DM is the ratio of the total solid content (g) to total weight of sample (g) times 100%.

Insoluble solids (IS) were measured after samples were centrifuged (20000 g for 30 min at room temperature, e.g. 23°C), obtained pellets were exposed overnight to a temperature of 80°C in an oven. The remaining was weighed. The % w/w IS is the ratio of insoluble solid content (g) to the total weight of the sample (g) times 100%.

Monosaccharide analysis by NMR on non hydrolysed samples

Quantification of fructose, glucose and sucrose in different carrot samples was performed by ¹ H NMR. Samples were freeze-dried and powdered if needed. The quantification is performed using targeted profiling (Chenomx).

1. Materials

1. Deuterated water (D ₂ 0, 99.96% D, DLM-6), Cambridge Isotope Laboratories.

2. 3-(Trimethylsilyl)propionic-2,2,3,3-c/4 acid, sodium salt (TSP, 98% D),Sigma- Aldrich.

3. (Difluorotrimethyl-silanyl-methyl)phosphonic acid (DFTMP), 153-RWH-196, Bridge Organics Co.

4. Ethylenediaminetetraacetic-c/i2-acid (EDTA-c/i2, 98% D), Isotec.

5. 0.2M phosphate buffer: 0.2M KH ₂ P0 ₄ Na2HP04 in D ₂ 0, pD 7.4 containing 0.1 % NaN ₃ .

6. Chemical Shift Indicator (CSI) solution: 10.32 mg of TSP (1 .2) and 2.22 mg of DFTMP (1 .3) in 30 ml of D ₂ 0 (1 .1 ).

7. EDTA-d _i2 solution: 50 mg of EDTA-d _i2 (1.5) in 10 ml of D ₂ 0 (1 .1 ).

8. 20 ml glass vials, pre-rinsed (to avoid formate and acetate contaminations).

2. Preparation of the sample solution

1. Sample solution. Add 3 ml of D ₂ 0 (1.1 ) to 30-50 mg (freeze dried) sample in a pre-rinsed disposable vial (1.8).

2. Mix the solution on a magnetic stirrer. 3. NMR sample solution. Transfer 600 μΙ of the sample solution, 100 μΙ of CSI solution (1 .6), 100 μΙ of EDTA-c/ _i2 solution (1.7), and 300 μΙ of phosphate buffer (1.5) into a 1 .5 ml Eppendorf tube and mix thoroughly.

4. Prepare a second (duplicate) NMR sample solution (2.1 , 2.2 and 2.3).

5. Centrifuge the NMR sample solutions (2.4) at 15000 x g for 10 minutes.

6. Transfer 650 μΙ of the clear NMR solutions into 5-mm NMR tubes for analysis.

3. NMR Measurements

1 D ¹ H NMR spectra were recorded with a NOESYGPPR1 D pulse sequence on a Bruker Avance III 600 NMR spectrometer, equipped with a 5-mm cryo probe.

The probe was tuned to detect 1 H resonances at 600.25 MHz. The internal probe temperature was set to 298K. 32 scans were collected in 57K data points with a relaxation delay of 10 seconds, an acquisition time of 4 seconds and a mixing time of 100 ms. Low power water suppression (16 Hz) was applied for 0.99 seconds. The data were processed in TOPSPIN software version 3.5 pi 1

(Bruker BioSpin GmbH, Rheinstetten, Germany). An exponential window function was applied to the free induction decay (FID) with a line-broadening factor of 0.15 Hz prior to the Fourier transformation. Manual phase correction and baseline correction was applied to all spectra. The spectra were referenced against the methyl signal of TSP (δ 0.0 ppm). The 1 D ¹ H NMR spectra were imported in Chenomx software (Chenomx NMR Suite Professional v7.63, Edmonton, Alberta, Canada). The relevant Chenomx models were fitted into the NMR signals of the target compounds, minimising the residual line. The reporting module calculates the compound concentration in the sample in three different units, i.e. %w/w, mg/g and mg.

Monosaccharide analysis by NMR on Saeman hydrolysed samples and the method for Saeman hydrolysis

The monomer composition of the polysaccharides in the preparation of carrot cell wall material was determined using a quantitative ¹ H-NMR method. The NMR quantification is done on samples which were freeze dried and powdered if needed and which were hydrolysed using the Saeman hydrolysis (Saeman, J.F., Moore, W.E., Mitchell, R.L and Millett, M.A., (1954) Tappi Journal, 37, 336-343). The N MR and the hydrolysis method is applied as described by de Souza et al. (De Souza, A.S., Rietkerk, T., Selin, C.G. and Lankhorst, P.P., (2013). "A robust and universal NMR method for the

compositional analysis of polysaccharides"; Carbohydrate polymers, 95, 657-663) with a few adaptations as detailed herein.

Approximately 15 mg of the sample was accurately weighed in a 15 mL glass culture tube. For pre-solubilization, 1 mL of 72% (w/w) D2SO4 in D2O was added to the sample. The sample was sealed and stirred at room temperature for 60 min. After this step, 6.2 mL D2O was added to the sample until the final concentration of 14% (w/w) D2SO4 in D2O was reached. The sample was sealed and incubated at 100 °C for 180 min. After the hydrolysis, the sample was allowed to cool down to room temperature.

Subsequently 1 mL of maleic acid internal standard solution was added. The final solution was pipetted into a 3 mm NMR tube. The ¹ H NMR spectra were recorded at 290 K with an Avance I I I 600 MHz NMR-spectrometer equipped with a 5 mm

cryoprobe. The NMR-spectra were recorded by using a standard pulse sequence (zg30). A relaxation delay of 60s was used. Below is a detailed procedure of the method. Reagents and standards:

Deuterated water (D20, 99.85% D), Eurisotop; Deuterated sulphuric acid (≥99%), Aldrich; D-(+)-Mannose (≥99%), Fluka; D-(+)-Glucose (≥99%), Sigma; D-(+)-Galactose (≥99%), Sigma; L-Rhamnose (≥99%), Aldrich; D-Galacturonic acid (97%), Sigma, L-(+)- Arabinose, (≥99%) Sigma, D-(+)-Xylose (≥99%) Sigma, Maleic Acid (99.78 ± 0.08 (g/g)%), Fluka.

Apparatus and equipment:

Apparatus: Avance I I I 600 MHz NMR spectrometer equipped with a 5mm-cryoprobe, (Bruker GmbH, Germany). (Multipoint) magnetic stirrer, type poly 15, 10 W,

(Variomag, USA). Magnetic stirrer with heater.Micro centrifuge, rotor 16 F24-1 1 (1 .5/2.0 mL), type 5416, (Eppendorf, Germany). Oven. Semi microbalance, readability 0.01 mg, Satorius CPA225D or equivalent.Magnetic stirrer, Ikamag RCT or equivalent. Vortex mixer, TM01 , LaboTech or equivalent Equipment: Disposable 16x100mm culture tubes with screwcap (Duran group, Germany). Disposable 1 .5 mL safe-lock PP micro-centrifuge tubes (Eppendorf, Germany). Disposable 3 mm NMR glass tubes (Bruker, Germany). 12 mm stirring bar (VWR, USA). Disposable 20 mL glass vials (Econo glass vials, Perkin Elmer, USA)

Sampling and preparation

(Note:Weigh all volumes accurately on a semi-microbalance).

Preparation of the polysaccharide sample solution (including solubilization): Weigh 15- 25 mg of each sample in a 15 mL glass tubes. Add 1 mL of 72 (w/w)%. D ₂ S0 ₄ in D ₂ 0 to each sample.

Solubilization: Add a stirring bar to each sample. Stir the samples at room temperature for 60 minutes using a multipoint stirrer (the glass tubes are placed in glass vials for stability). After the solubulization step, add 6.2 mL D ₂ 0 to each sample in order to obtain a final D ₂ S0 ₄ concentration of 14 (w/w) % in D ₂ 0.

Hydrolysis: Incubate the samples in an oven for 90 (180 for pectin containing samples) min at 100 °C. After 90 (180) min, let the samples cool down to room temperature. NMR Sample preparation: Carefully add 1 mL of maleic acid internal standard solution (0.5 mg/ml in D ₂ 0) and stir the solutions at a magnetic stirrer for 5 minutes. The NMR samples were transferred to 3-mm NMR tubes for analysis.

Preparation of the sugar recovery standard (SRS) solutions:

For every individual monosaccharide, dissolve 50 mg of the monosaccharide (Mannose, D-(+)-Glucose, D-(+)-Galactose, L-Rhamnose, Xylose, Arabinose or Galacturonic acid ) in 20 mL D ₂ 0 in a 20 mL glass container. For each monosaccharide, transfer four aliquots of 3.1 mL from this solution to 15 mL glass tubes. Slowly add 500 μί 72 (w/w) % D ₂ S0 ₄ in D ₂ 0 to each sample. Stir the solutions at a magnetic stirrer for 5 minutes. Hydrolysis:

Incubate half of the samples in an oven for 180 min at 100 °C. After this hydrolysis step, let the samples cool down to room temperature. The remaining SRS samples should not be hydrolyzed and thus kept out of the oven.

NMR Sample preparation: 500 μΙ of maleic acid internal standard solution was added to the samples. The solutions were stirred for 5 minutes on a magnetic stirrer. The NMR samples were transferred to 3-mm NMR tubes for analysis.

NMR data processing

Process the data in TOPSPIN software version 3.1 pi 5 (Bruker BioSpin GmbH, Rheinstetten, Germany). Apply an exponential window function to the free induction decay (FID) with a line-broadening factor of 0.3 Hz prior the Fourier Transformation (FT). Apply manual phase correction, baseline correction and chemical shift calibration to the FT NMR spectrum.

For this data set, the spectral processing was done with PERCH NMR software:

A small library was created for the deconvolution of galacturonic acid, galactose, arabinose, glucose, xylose, mannose and rhamnose. Each of the compounds was characterized by a set of spin particles and parameters for chemical shifts, coupling constants, line widths and intensities (populations). These parameters were obtained by fitting experimental spectra of the pure model compounds recorded in D20 using the PERCH NMR software. All parameters were iteratively optimized. The internal standard (IS) consisting of maleic acid was used to calculate the absolute compound concentrations from the fitted populations. The spin particles of the internal standard were included in each fit.

Calculation and expression of result Calculation for the internal standard method:

Calculation of the degradation of a monosaccharide (DM) after hydrolysis (SRS- control). This should be done for each monosaccharide present in a polysaccharide.

MS

D M

P M, H

Where I = measured integral

N = number of protons

MW = molecular weight (g/mol)

W = actual weight (mg)

V = total volume of solution (mg)

f = used fraction the solution (mg)

P = purity (%)

And IS = internal standard

MH = monosaccharide hydrolysed

MS = monosaccharide sugar recovery standard (non-hydrolysed)

Calculation of the monosaccharide concentration in a polysaccharide. This should be done for each monosaccharide present in this polysaccharide.

fl _MP N _IS MW _MP W _IS

^MMPP — = ^UD MM -' I[ -T

s N M, P MW _JS W, IS

MP

Where

MP = monosaccharide present in polysaccharide

Calculation of ratio Galacturonic acid to bound Glucose

The GalA Glu ratio is calculated from the galacturonic acid amount in the sample after Saemen hydrolysis [GalA] divided by the total amount of Glucose in the sample

[Glucose total after hydrolysis ] which is corrected for the free glucose.

Free glucose (free Glu) analysed is the sum of the monomeric glucose in the sample without hydrolysis and the glucose content in the quantified sucrose in the sample without hydrolysis. The glucose amount in sucrose is 50%. In formula:

Galacturonic acid / Glucose (bound) =

[Galacturonic Acid] / ( [Glucose total]- ([Glucose monomeric ]+ [GlUCOSe from sucrose] ))

In which: Glucose total= Glucose after hydrolysis

Glucose monomeric=Glucose monomeric determined for the non-hydrolysed sample

Glucose from sucrose= The glucose content in sucrose as determined for the non-hydrolysed sample For the calculation of the ratio of Galactose or Arabinose/ Glucose (total giucose-free glucose) a similar equation was used as above but instead of [GalA] the concentration of galactose or Arabinose was used respectively.

Scanning confocal laser microscopy (CSLM).1 ml of sample was stained with 1 droplet Congo Red solution of 1w% and gently stirred through the sample. Subsequently, the dyed samples are kept at room temperature for approximately one hour to reach equilibrium and to properly stain the whole sample. The stained sample is placed on a glass slide. Congo red was used to visualize fiber dispersions because of its strong affinity with cellulose. Imaging is done using the Leica TCS-SP5 confocal microscope with the DMI6000 inverted microscope. A solid state laser emitting at 561 nm was used for excitation. The emission bandwidth used was 565 nm to 677 nm.

Silverson shear treatment

For the shear treatment using a Silverson L4RT homogenizer the laboratory workheads used were:

Workhead 1 ) The Square Hole High Shear Screen™= square hole screen, used as pre-treatment for 5-10min at 3000-5000rpm.

Workhead 2) Fine Emulsor Screen (smallest hole screen), a workhead with 2 mm diameter round holes used for 10 min. between 5000-7700 rpm. This step can be used multiple times.

Workhead 3) Slotted Disintegrating Head, this is a slit shaped holes head. Standard Silverson redisoersion treatment

The Standard Silverson redispersion treatment used herein was was performed by combining the amounts of a sample of a preparation of carrot cell wall material and demineralised water required to yield 500 grams of a dispersion comprising 1.00 wt% by weight of dry matter of the carrot cell wall material to the total weight of the composition. The combined mixture treated with a Silverson L4RT homogenizer for 10 min at 6000 rpm using workhead 2 (fine emulsor screen) and a 1 L laboratory beaker.

EXAMPLES

Examples 1 , 2 and comparative example A: Preparation of carrot cell wall material from fresh carrot

Carrots from a local supermarket were peeled and cut in approximatively 1 -2 cm slices. After washing, the slices were heated for 30 minutes in water at 95°C. Then after cooling, the pieces were mechanically disrupted in (one part of) demineralized water using a blender (Kenwood Chef Glass liquidiser) at the maximum speed for 1 to 3 min. The resulting carrot puree was diluted in water to a total solids content of 1 .7 wt-% (Comparative Ex. A). To remove water-soluble solids from the fresh vegetable dispersion, water was added to the undiluted puree (of comparative A) and the resulting slurry was centrifuged for 30 minutes at 4500 rpm at room temperature, yielding a pellet of cell wall material. The operation was repeated by redispersing the pellet and centrifuging until a Brix value of the supernatant of < 0.2° was obtained. The final washed and centrifuged pellet was diluted in water using a Silverson mixer for 2 minutes at 2000 rpm to a total solids content of 1 .7% (Ex. 1 ). For the preparation of Example 2, a slurry of carrot cell wall material, prepared according to Ex. 1 , was mechanically processed using an overhead mixer (Silverson L4RT-A mixer, workhead 2, fine emulsor screen with 2 mm diameter round holes during 10 min. at 5000 rpm) and subsequently diluting the slurry to 1.0% total solids content (Ex. 2).

The analysis of the samples presented in Table 1 demonstrates that preparations of carrot cell wall material prepared according to the present invention can be used to structure aqueous phases as evidenced by the high shear storage moduli of Examples 1 and 2. In contrast, the untreated cell wall material of Comparative Example A does not provide any structuring. Table 1. Properties of carrot slurry

^* The Total solids (TS) were determined by drying a known amount of the carrot puree (comparative A) or the pellet (Ex. 1 ) in an oven at 105°C, during 16h under 100 mbar vacuum. The TS was determined on an average of three measurements.

^** The shear storage modulus G' after 5 min equilibration was determined using the method defined hereinabove.

Example 3. Preparation from frozen carrot press cake

Frozen carrot press cake (500g) from a carrot juice process (ex Van Rijsingen) was mixed with 3500g demineralised water and heated to boiling temperature in a 5L pot on an induction stove. After 30 min heating at T> 95°C under stirring with a fork, the mixture was cooled down and demineralized water was added to compensate for evaporation. The above process was repeated with a second batch and the two batches were combined. The resulting slurry was centrifuged in a Sigma 8K centrifuge using 6X1.5L PE buckets during 15 minutes at 5100 rpm (=8450g). The resulting pellet was resuspended in demi water in a total volume of 8800ml and centrifuged again. After the second washing cycle the pellet was resuspended in a total volume of 8800 mL and homogenised using a Silverson BX mixer with a slotted disintegrator work head (workhead 3) in a 10L plastic vessel for 5 minutes at 3000 rpm. The homogenized slurry was washed two more times with a centrifugation time of 20 minutes at 8450g. The Brix value of the supernatant of the final washing was <0.1 °.

The resulting cell wall material was split in 400 g batches and different batches were redispersed in demineralised water at a concentration of 0.8, 1.2 and 1.7 % DM, respectively. The batches were redispersed with a Silverson mixer (with 2 mm square hole screen, workhead 1 ) for 8 min at 8000rpm (Ex. 3). The results summarised in Table 2 demonstrate that good structuring can also be obtained with a preparation of carrot cell wall material according to the invention that was prepared using frozen carrot press cake as the starting material. Table 2 Ex 3 storage modulus as function of concentration and homogenization method

^* The shear storage modulus G' after 5 min equilibration was determined using the method defined hereinabove.

Example 4. Preparation of carrot cwm with repeated washing

Frozen carrot press cake (375g ex Van Rijsingen) was allowed to thaw and 3000 g of boiled demineralised water was added. The suspension was heated in the microwave for 5min at 1000W and subsequently blended using a Thermomix (30min at 90 °C, speed 4, ex Vorwerk, from Thermomix Benelux NV CNUDDE)). The suspension was then centrifuged using a Beckman Coulter Avanti J-26S XP centrifuge, with 6x 400g in centrifuge buckets, for 30rminut.es at 10000g. The resulting pellet was re-dispersed in demi water to a total volume of 1600 mL and centrifuged another time to remove more soluble compounds. At each washing step, water was added to reach a similar weight and a sample was taken for DM analysis, conductivity and rheology measurements. In total 4 washing steps were done and after each washing step a sample was taken. Also a sample was taken for the non-washed condition. For the rheological measurements the %DM of carrot cwm of the different suspensions was adjusted to 1 % by addition of demineralized water. The samples were then homogenised for 10min at 6000rpm using a bench top Silverson mixer (using workhead 2, fine emulsor screen). Results of the rheology of the dispersions of carrot fibers in water and conductivity measurements are presented in Table 3. Table 3. Conductivity and storage modulus G' at 1 % DM carrot cell wall suspension in demineralised water after different washing conditions and Silverson treatment 10min 6000rpm.

Example 5 fractionated carrot cwm

154g of carrot press cake was mixed into boiled demineralised water. The suspension with a total weight of 1 .5 kg was heated in a Microwave for 5min at 1000W, then heated and blended in Thermomix during 1 h at 90°C, then Silverson treated for 5 min 5000 rpm (using workhead 1 , square hole screen) followed by 10 min Silverson (using workhead 2, fine emulsor screen) treatment at 77000 rpm, then frozen at -80°C and finally freeze dried.

The freeze-dried powder was rehydrated in demineralized water at 1 .50 % DM of carrot cwm and blended in a Waring blender for 5 min at speed 3 and 100 ml was washed using filtration on Miracloth (pore size 25 μηη, from CalBiochem). A small sample was taken to measure %DM and to determine the yield. The residue was found to contain 75% of the DM of the filtered carrot cwm. The residue was mixed in demineralized water at a DM content of 1 .5% and the carrot cwm samples (before and after filtration) storage moduli were measured after 24 h storage (Table 4).

Table 4: Storage moduli of different particle size fractions of Ex. 5

Ex. 5 rehydrated, mixed in Waring blender

G' eq. 5min [Pa] ±SD G' str. sw [Pa] ±SD

before filtration 1329 ± 153 68 ± 2

After filtration 7645 ± 563 326 ± 13 The strong increase in G' of the larger particle size indicates that the larger sized particles >25μηι have a stronger contribution to the rheology than the particles <25 μηι.

Examples 6-9 & comparative B,C.

In order to determine the monosaccharide composition of preparations of carrot cell wall material.

Example B was made by taking 50 g FW (fresh weight) of fine cut raw carrot (from van Rijsingen, stored frozen) and drying the frozen carrot using freeze drying. The freeze dried carrot was milled to a powder using a Waring Laboratory Blender for 2 min at speed 4.

Example C was made by taking 50 g FW (fresh weight) of fine cut raw carrot (from van Rijsingen, stored frozen), letting it thaw for 15 min at room temperature and then washing the carrot at room temperature on a Miracloth filter (poresize 25 μηη, from CalBiochem) with an excess of demineralised water until the filtrate was almost colourless. The residue was collected and freeze dried and then milled to a powder using a Waring Laboratory Blender for 2 min at speed 4.

Example 6: 154g of fine cut raw carrot (from van Rijsingen) was used which was mixed with 1.5L of boiled demineralised water and heated in the microwave for 5min at 1000W. The suspension was then blended using a Thermomix (30min at 90 °C speed 4). The suspension was then washed using a Buchner funnel and a Miracloth filter (25μηι pore size, from CalBiochem). Demineralised water (around 4L) was used for washing the fibers until the filtrate was almost transparent. After being washed, the residue left on the filter was collected and freeze dried and then milled to a powder using a Waring Laboratory Blender for 2 min at speed 4.

Example 7 was made similar to Example 6, but after the washing step the carrot residue was collected and demineralised water was added to prepare a suspension of 1 % DM. Then the carrot suspension was treated by shear using a bench top Silverson (L4RT) for 10 min 3000rpm using the square hole screen workhead 1 (large pores) followed by 10 min 6000 rpm using the fine emulsor screen workhead 2. The suspension was then washed using a Buchner funnel and a Miracloth filter (25μηι pore size, from CalBiochem). Demineralised water (around 4L) was used for washing the fibers until the filtrate was almost transparent. After being washed, the residue left on the filter was collected and freeze dried and then milled to a powder using a Waring Laboratory Blender for 2 min at speed 4.

Example 8 was made similar as described for Example 7 except that after the Silverson treatment a HPH treatment was applied using the lab scale Panda Plus High Pressure Homogeniser (from Niro Soavi) between 600-1000 bar. Then the HPH-treated carrot suspension was frozen at -80 °C, then freeze dried, then milled to a powder using a Waring Laboratory Blender for 2 min at speed 4.

Example 9: To 125g of fine cut raw carrot (from van Rijsingen) 625 g boiled

demineralised water was added and heated in the microwave for 5min at 1000W. Subsequently the suspension was blended using a Thermomix (30min at 90 °C speed 4). Washing of the fibers was carried out using a Buchner funnel and a Miracloth filter (25μηι pore size). Demineralised water (~ 4L) was used for washing the fibers until the filtrate was almost transparent. The washed residue on the filter was collected and demineralised water was added to reach a total weight of 1 .3kg. This suspension was homogenized with a Silverson mixer (L4RT) firstly during 5 minutes at 3000 rpm (using workhead 1 , square hole screen pores) followed by 10 min 7700 rpm (using workhead 2, fine emulsor screen). Subsequently the homogenised carrot preparation was heated in the Thermomix during 30min at 90 °C (speed 4) and homogenized with a Silverson mixer (L4RT) for 10 min at 7700 rpm (using workhead 2, fine emulsor screen) and then washed (filtration 25 micro meter pore size) and homogenised in a Niro Soavi High Pressure Homogeniser at 1000-1600 bar. Washed again using Miracloth and again a HPH treatment was applied under similar conditions as described above. Then the 2xHPH treated carrot suspension was frozen at -80 °C, then freeze dried, then milled to a powder using a Waring Laboratory Blender for 2 min at speed 4.

The preparation of the Examples is summarised in Table 5.

Table 5. Unit operations for the preparation of Ex. 5-8 and comparative F and G.

Ex Sample name Starting wash Heating+ homogenisation material blending

B raw Freeze Raw carrot

Dried carrot c frozen carrot Pureed Demi

washed (room carrot water

temp. ) FD

6 Thermomix Heat + Pureed Demi 30' @

blended/ washed parrot water 90C

FD

7 Washed and Pureed Demi 30' @ Silverson 10' sheared carrot water 90C @ 6000 rpm

(Silverson)

washed FD

8 processed FD Pureed Demi 30' @ HPH

"IxHPH carrot water 90C

9 processed FD 2x Pureed Demi 30' @ 2ΉΡΗ (with

HPH (washed in carrot water 90C wash in between) between)

Carbohydrate composition of freeze dried carrot samples was determined after Seaman hydrolysis by NMR (Table 6).

Table 6 Monosaccharide analysis by NMR after Saeman hydrolysis of freeze dried

(FD) preparations of carrot cell wall material

^* (GalA = Galacturonic acid; Gal = Galactose; Ara = Arabinose; Glu = glucose) ^** Glucose content was determined as bound glucose in the insoluble fraction of the carrot cell wall material as described in the Method section

Free glucose content in sample F and G were respectively 10.58 and 9.46% (w/w), whereas the free glucose content in the other samples was below 0.01 % (w/w).

Examples according to the invention have a GalA:Glu ratio below 1 .0, whereas the comparative examples that did not undergo essential unit operations have ratios below 1.0. Example 10 and Comparative Examples D-G.

Examples 10 and Comparative Examples D to G were performed to assess the effect of hydrogen peroxide bleaching on the monomer composition of carrot cell wall material. Pieces of carrot tissue were mixed with hot water to a dry matter content of 5% (w/v). The mixture was heated in the microwave to reach 90C. Subsequently the suspension was blended using a Thermomix (120 min at 90 °C speed 3-4). Then Silverson treatment for 5 min using workhead 1 , square hole screen pores) at 6000 rpm followed by 10 min at 7700 rpm (using workhead 2, fine emulsor screen). The resulting slurry was filtered through a 0.500 mm filter and incubated with 17.5% and 0.175% H2O2 during 1 , 4 and 6 hours, respectively, at 90°C and subsequently washed using a 0.25mm sieve and with excess of demineralised water and homogenised according to the scheme in Table 7. Example 10 according to the invention was incubated as Ex. G, but in the absence of H2O2. The samples were freeze-died prior to further analysis. The monosaccharide composition of the freeze dried samples was determined after Seaman hydrolysis by NMR as described in the Method section. (Table 8). β-carotene content was also determined by NMR.

Table 7: Overview of process details and yield of differently processed carrot samples

Process

2nd homo- Yield

Ex: heat H2O2 heating genisation sieve pH % % t [min], Pore size Before dry t,T (w/w) t,T speed [rpm] [mm] washing wt.

10 2h,90C 0 6h,90C 15 min, 8000 0.25 4.7 88

D 2h,90C 17.5 1 h,90C 15 min, 8000 0.25 3.24 47

E 2h,90C 17.5 3.5h,90C 15 min, 8000 0.25 2.47 13

F 2h,90C 0.175 1 h,90C 0.25 4.67 61

G 2h,90C 0.175 6h,90C 0.25 3.1 48

Table 8 Monosaccharide analysis by NMR after Saeman hydrolysis of freeze dried (FD) preparations of carrot cell wall material

^* (GalA = Galacturonic acid; Gal = Galactose; Ara = Arabinose; Glu = glucose) * ^* Glucose content was determined as bound glucose in the insoluble fraction of the carrot cell wall material as described in the Method section. Free glucose content in sample F and G were respectively 10.58 and 9.46% (w/w), whereas the free glucose content in the other samples was below 0.01 % (w/w).

The GahGlu ratios of the comparative sample treated with H2O2 are below 0.15 whereas the GahGlu ratios of the Examples according to the invention are above 0.15. The residues after washing of the cwm of example 9 and of the comparative examples D-G were collected and suspended to a 1 % cell wall material (cwm). Physical properties (rheology, yield, pH) were measured (Table 9). Table 9.

sweep ^** and viscosity was given at 0.1 at 1 and 10/s and viscosity was presented using the Herschel-Bulkley equation (all equation data are given). % Yield is based on 100 % cwm in starting material.

The yield and all rheological values of example 4 are higher than the comparative example which are treated by H2O2.

Examples 11 -16. Preparation of carrot cwm using different order of unit operations)

Preparations of carrot cell wall material were prepared using the unit operations heating, blending, washing and mechanical shearing as described in Ex. 1 , but applied in different order as presented in Table 10. When the heating step was not the first one, carrot slices were blanched to inactivate enzymes and mechanically disrupted in water (ratio carrot/water: 1/1 ) using the Kenwood blender at maximum speed for 2 min. Dispersions were diluted to 1.0% total solid content before determination of the storage modulud G'.

Table 10. Sample treatment and storage modulus of Ex. 11 -16

^* The shear storage modulus G' after 5 min equilibration was determined using the method descibed above. Examples 17-21 and comparative example H . Preparation of carrot cwm using different combinations of unit operations

To 125g of carrot press cake 625 g boiled demineralised water was added and heated in the microwave for 5min at 1000W. Subsequently the suspension was blended using a Thermomix during 30min at 90 °C speed 4 (Comparative example H). Washing of the fibers was carried out using a Buchner funnel and a Miracloth filter (25μηι pore size). Demineralised water (~ 4L) was used for washing the fibers until the filtrate was almost transparent. The washed residue on the filter was collected and demineralised water was added to reach a total weight of 1.3kg (example 17 (H+W) particle size of cwm is larger than 25μη"ΐ). An aliquot of 300g of this suspension was homogenized with a Silverson mixer (L4RT) firstly during 5 minutes at 3000 rpm using a ring with large pores (workhead 1 , square hole screen) and subsequently during 10min at 7700rpm using a ring with small pores (workhead 2, emulsor screen) (Example 18; H+W+S). To prepare example 19 (H+W+S+H+S), example 18 (H+S+W) was heated in the

Thermomix during 30min at 90 °C (speed 4) and subsequently homogenized with a Silverson mixer (L4RT) for 10 min at 7700 rpm (workhead 2, emulsor screen).

Example 19 (H+W+S+H+S) was freeze dried and resuspended in water and homogenized with a Silverson (L4RT) for 10 min at 6000 rpm (workhead 2, emulsor screen) to obtain example 20 (H+W+S+H+S+FD).

To obtain example 21 (H+W+S+H+S+W+S+HPH), sample 19 (H+W+S+H+S) was washed using Miracloth filter (25μηι pore size) and the resuspended pellet was homogenized with a Silverson mixer (L4RT) for 10 min at 7700 rpm (workhead 2, emulsor screen) followed by the shear treatment in a High Pressure Homogeniser (HPH) using a lab scale Panda Plus HPH (from Niro Soavi) between 600-1600 bar. CSLM images of Figures 1 -3 show that all samples contained recognizable cell walls, including intact cell wall enclosures and cotyloid cell wall fragments and that increased shear leads to smaller particle size. It was found that the majority of the particles had a size below 500 μηη as not more than 2% by dry weight of the cell wall material was retained on a 500 μηη sieve (woven wire stainless steel sieve from Endecotts).

The self-suspending capacity of the above carrot cell wall preparations was determined in a graduated cylinder of a 0.3% (w/v) on dry weight basis in demineralized water. As a control the starting carrot material was diluted to a content of 0.3% insoluble cell wall material in demineralized water.

Table 11. Suspending properties and storage modulus of carrot cell wall preparations

measured by the rheology method used. Example 22 - Reversible drying

The Silverson-homogenized carrot cwm from Ex. 3a was centrifuged and the pellet was mixed with a solution of MaltoDextrin 20 (MD20) 50% w/w in a 10L stainless steel bucket to obtain a 1 :1 ratio of cwm:MD20 on a dry matter base. The mixture was centrifuged and the pellet was spread on metal trays and cooled to -20 a -30°C using a blast freezer within 1 hr and were subsequently dried at -20°C in a freeze dryer. The cake was broken into 1 -2 cm pieces (manually) and filled into the blender reservoir till half volume. The cake pieces were pulverised using a blender using short bursts at low speed counting up to 10-15 s. The powder (light orange coloured) was collected in PE containers and stored at ambient temperature in the dark. The dried cwm was re- dispersed at a 1 % dry matter (fibre) content (w/w) in demi water. 100 ml of the suspension is placed in a 100 ml graduated cylinder and the volume occupied by the cwm material was determined. Table 12 Suspension properties of carrot cwm of example 22

Example 23 Dried carrot cwm; solvent exchange

An alcohol insoluble residue (AIR) was made from the carrot cwm preparation 19 (H+W+S+H+S) by ethanol washing. To this end an aliquot of example 19

(H+W+S+H+S) was added to 96% ethanol in ratio 24/76 carrot cwm in water/ethanol respectively and then washed with 2 volumes of 96 (v/v) ethanol on a Whatmann (No 1 13) filter and subsequently with 96% ethanol. The residue on the filter was collected and dried in the oven overnight at 80°C. Then this AIR powder was rehydrated to form suspensions containing 1 %DM of cwm which were sheared by Silverson mixer for 10 min at 6000 rpm (using workhead 2, emulsor screen) (Ex. 23a) and for an additional 10 min at 7000 rpm (using workhead 2, emulsor screen) (Ex. 23b).

All differently processed suspensions were adjusted to similar pH prior to the rheology measurements of which showed the solvent exchange treatment (preparation of drying the Alcohol Insoluble solids) did not influence the storage modulus G' of the carrot cwm dispersion when rehydrated in water.

Table 13. Storage moduli of carrot preparations Ex 19 and Ex 23a and b.

Examples 24-25. Sauce structured with Carrot cwm

The preparation of carrot cell wall material of Example 2 was used to structure a tomato-based sauce. For Ex. 24 and Ex 25, 30% and 70% (respectively) of the tomato paste (28-30 °Brix) was replaced by 52 and 122 g of a 5.0% (DM) the preparation of cwm, respectively, to obtain 0.52 and 1 .22% of carrot cwm material content on a dry weight basis (Table 14). The sauce was prepared by mixing the ingredients with a household mixer. Rheological and other physical properties of the sauces were determined (Table 15).

Table 14. Composition of tomato-based sauce

Reference Ex. 24 Ex. 25

ingredient G g g

Tomato dice 52.7 52.7 52.7

Tomato 174 122 52

paste

Water 206 191 171

Tomato pulp 53 53 53

sugar 10 25 45

Salt 4 4 4

Carrot cwm 0 52 122 Table 15. Physicochemical properties tomato-based sauce

Example 26. Oil in water emulsion containing carrot cwm

The preparation of carrot cell wall material from Ex. 3a was used to structure an oil in water emulsion. The dried preparation was rehydrated in water and mixed with the egg for 30 seconds. After 20 sec the water phase was added and mixed. Then oil was added and the mixture was emulsified in Fryma Delmix.

Table 16. Composition of o/w emulsion

Table 17. Stevens value and syneresis of o/w emulsion

Stevens Syneresis pH

7 days 0 days 7 14 days

days

Ex. 26 121 .5 0 0.02 0.49 3.71