DOI: 10.3303/CET23102026 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Paper Received: 22 February 2023; Revised: 18 April 2023; Accepted: 22 August 2023 
Please cite this article as: Pirozzi A., Donsì F., 2023, Structuring Vegetable Oils Through the Formation of Capillary Suspensions: Comparison 
of Wheat Middlings and Pure Cellulose Processed by High-pressure Homogenization, Chemical Engineering Transactions, 102, 151-156  
DOI:10.3303/CET23102026 
  

 CHEMICAL ENGINEERING TRANSACTIONS  
 

VOL. 102, 2023 

A publication of 

 

The Italian Association 
of Chemical Engineering 
Online at www.cetjournal.it 

Guest Editors: Laura Piazza, Mauro Moresi, Francesco Donsì 

Copyright © 2023, AIDIC Servizi S.r.l. 

ISBN 979-12-81206-01-4; ISSN 2283-9216 

Structuring Vegetable Oils Through the Formation of Capillary 

Suspensions: Comparison of Wheat Middlings and Pure 

Cellulose Processed by High-Pressure Homogenization 

Annachiara Pirozzi, Franceso Donsì* 

Department of Industrial Engineering, University of Salerno, via Giovanni Paolo II, 132, 84084 - Fisciano, Italy 

fdonsi@unisa.it 

Reducing the intake of harmful trans-fats and saturated fats in the diet, by replacing detrimental fats with 

healthier oils, without affecting the organoleptic properties of the food product, represent a formidable challenge 

for the scientific community. In this scenario, this work explores a possible strategy for structuring sunflower oil 

by investigating the formation of capillary suspensions using wheat middlings (WM) and pure cellulose (CL) as 

a structuring solid fraction. High-pressure homogenization (HPH), a purely mechanical cell disruption 

technology, was directly applied to oil suspensions of WM or CL. Subsequently, the addition under high-shear 

mixing (HSM) of different amounts of an immiscible secondary fluid, water, to the oil suspensions, led to WM 

and CL particles bridging and network formation, through the development of attractive capillary forces among 

the particles. The effect of water and particles characteristics on the rheological behavior of the oil suspensions 

was investigated. The presence of water caused initially an increase in viscosity and then a decrease, as water 

concentration exceeded a critical value, with an inversion from a continuous oil phase to a continuous aqueous 

phase. Moreover, the oxidative stability of the capillary suspensions was evaluated, during accelerated aging. 

The proposed approach not only does not suffer the presence of water, but significantly improves the oxidation 

stability with respect to the pure oil 

1. Introduction 

In the last decade, oil structuring has gathered increasing interest because of its potential use in the food, 

pharmaceutical, and biotechnology industries for replacing unhealthy saturated fats. More specifically, oleogels 

have been proposed as novel systems in food preparations incorporating vegetable oils, to enhance their 

techno-functional properties and nutritional value, without penalizing taste and mouthfeel (Le Révérend et al., 

2010). Oleogels are systems characterized by a continuous liquid phase physically immobilized by a self-

assembled network of gelators, such as waxes, monoglycerides, fatty acids, plant sterols, and esters. A heating 

step is generally applied to melt the gelator in the liquid oil phase under gentle stirring for its full dissolution. By 

cooling below the gelation transition temperature, the gelator self-assembles into a thermoreversible three-

dimensional network that entraps the surrounding oil phase (Blake et al., 2018; Zhao et al., 2021). Oleogelators 

should typically feature (i) the presence of lipophilic and interacting parts, (ii) surface activity, and (iii) 

thermoreversible characteristics. However, for food use, the gelators should be of natural origin and considered 

as GRAS, hence severely reducing the available options. For this reason, capillary suspensions have emerged 

as a promising potential alternative for oil structuring. The addition of small amounts of a secondary immiscible 

fluid (generally water) to a suspension (such as particles in oil) triggers attractive capillary forces, leading to 

particle bridging and the formation of a 3D network (Koos, 2014; Mustafa et al., 2018). In this way, a transition 

from a fluid-like to a gel-like state is promoted. The rheological properties can be tuned by controlling the 

attractive capillary forces (e.g. particle size or water fraction), with an increase in yield stress and shear viscosity 

by several orders of magnitude (Koos and Willenbacher, 2011), increasing also suspension stability. Depending 

on how primary and secondary fluid wet the particles, capillary suspensions can be classified in the pendular 

state, spherical agglomerates, bi-continuous, Pickering emulsion, and capillary state (Bossler et al., 2016). 

151

mailto:fdonsi@unisa.it


This work aims at structuring sunflower oil by high-pressure homogenization treatment applied to wheat 

middlings (WM) or cellulose (CL) particles directly suspended in oil, with the final goal of obtaining a gel-like 

product without the use of unhealthy fats or undesired additives, but only with the use of water as secondary 

fluid. The use of WM enables also the reduction of the caloric content of the oil, while sustainably valorizing an 

agri-food residue (AFR) of the wheat milling process, hence contributing to implementing the circular economy 

across the food chain. Moreover, the antioxidant content of WM could also contribute to slowing down oil 

oxidation processes. The use of pure cellulose, which is especially challenging when considering the hydrophilic 

nature of the sugar monosaccharides building blocks, enables the exploitation of a natural biopolymer with high 

biocompatibility, ecological acceptability, renewability, and low toxicity. By advancing the understanding of the 

mechanism of network formation of oil suspensions using high-pressure homogenization (HPH), a contribution 

will be given to developing new applications in oil structuring. 

2. Materials and methods 

2.1 Raw materials 

Wheat middlings (WM) was kindly supplied by a local milling industry (Albanella, Salerno, Italy), and stored at 

room temperature until used. Commercial cellulose (CL) Arbocel® BWW40, with a declared average length of 

the cellulose fibers of 200 µm, was kindly supplied by Rettenmaier Italia (Brescia, Italy). Sunflower oil was 

purchased from a local market (Olio di Semi di Girasole Basso, San Michele di Serino (AV), Italy). Its 

composition (on a weight basis) comprises 10.3% wt of saturated fatty acids, 19.5% wt of monounsaturated 

fatty acids, and 65.7% wt of polyunsaturated fatty acids. All water used throughout this work was purified by the 

Milli-Q water purification system (Barnstead™ Pacific TII Water, Thermo Scientific, Waltham, MA, USA). 

2.2 Formulation of oil suspensions 

WM was suspended in sunflower oil at 30% w/w and cellulose fibers at 5%, which correspond to the cellulose 

content in WM particles (data not shown). The solid particles were firstly dispersed using a high-shear mixer 

(HSM) (Ultra Turrax T25, IKA Labortechnik, Germany) at 20,000 rpm for 5 min within an ice bath. The obtained 

suspensions were then subjected to high-pressure homogenization (HPH) treatment (80 MPa for 20 equivalent 

passes) using an in-house developed system, using cooling water at 10 °C. 

2.3 Characterization of oil suspensions 

The particle size distribution of the samples micronized in oil by HSM or HPH was measured by laser diffraction 

using a Malvern Mastersizer 2000 (Malvern Instruments Ltd., UK). The oil phase of oil dispersions was 

substituted by water by washing them three times with saponin from quillaja bark (Sigma-Aldrich, St. Louis, MO, 

USA)) at 0.5% w/v in water, followed by centrifugation at 6,500 rpm. The diameters corresponding to the 10th, 

50th, and 90th percentile of the cumulative distribution (d(0.1), d(0.5), and d(0.9), respectively), the volume-based 

and surface-based mean diameter (D[4,3] and D[3,2], respectively), were calculated as previously described 

(Pirozzi et al., 2021). The mean value of three replicates was determined. 

2.4 Capillary suspensions formulation 

The capillary suspensions were obtained by adding a secondary fluid (distilled water) at 0.05, 0.30, and 0.50 

values of saturation ratio SR (Equation 1) to the HPH-treated oil suspensions under HSM (20,000 rpm for 10 

min) within an ice bath (Bossler et al., 2017). 

SR =
𝑉𝑤𝑎𝑡𝑒𝑟

𝑉𝑜𝑖𝑙 + 𝑉𝑤𝑎𝑡𝑒𝑟
 (1) 

2.5 Capillary suspensions characterization 

The suspension morphologies were observed with an optical microscope (Nikon Eclipse TE 2000S, Nikon 

Instruments Europe B.V., Amsterdam, The Netherlands), coupled to a DS Camera Control Unit (DS-5M-L1, 

Nikon Instruments Europe B.V, Amsterdam, The Netherlands) for image acquisition and analysis. 

Rheological analyses of capillary suspensions were performed in a rotational rheometer (AR 2000 rheometer, 

TA instruments, Newcastle, DE, USA), equipped with a concentric cylinder (15 mm stator inner diameter, 28 

mm rotor outer diameter, 42 mm cylinder immersed height, 2° cone angle) and plate-cone geometry (40 mm 

diameter, 2° cone angle and 1 mm fixed gap width). The apparent yield stress was evaluated using a shear rate 

ramping. Flow curves were obtained by continuously varying the shear rate from 0.1 s-1 up to a rate of 200 s-1 

at 20 °C. Measurements were repeated five times on two independently prepared samples and were reported 

as means ± standard deviations for apparent yield stress. 

152



2.6 Accelerated oxidation stability test 

The primary oxidation compounds analysis is always carried out by the peroxide value (PV) to assess the quality 

of oil at its stability. For the Schaal oven accelerated stability test, approximately 1 g of each sample was put in 

vials and stored in an oven at 65 ± 2 °C in absence of light after 15-days of storage (Lužaić et al., 2022). The 

PV value was determined iodometrically (visually) with a starch indicator and a sodium thiosulfate standard 

solution according to ISO 3960:2007 (International Organization for Standardization). 

2.7 Statistical analysis 

All experiments and analyses were performed in triplicate unless differently specified and the results were 

reported as means ± standard deviations. Statistically significant differences among mean values were analyzed 

by one-way variance (ANOVA and Tukey test (p < 0.05), performed with SPSS 20 (SPSS IBM., Chicago, USA) 

statistical package. 

3. Results and discussions 

3.1 Oil suspensions 

The size distribution is a crucial analysis that affects the physical properties of oil suspensions. The particle size 

distributions for WM and CL-based oil suspensions treated by HSM and HPH (Figure 1) give information on the 

effect of the two different mechanical treatments. For the WM-based oil suspensions treated by HPH, the 

distribution was narrower and more uniform than for HSM-treated samples, as already observed in the case of 

tomato peels, because of the disruption of individual plant cells into fibrous debris (Jurić et al., 2019). In contrast, 

HPH treatment applied on cellulose-based oil distribution induced only a slight decrease in the particle size with 

respect to the HSM-treated samples, likely because of the initial finer size of the used cellulose, limiting the 

capability of HPH treatment to further break down the cellulose fibrils. 

Particle size ( m)

1 10 100 1000 10000

V
o

lu
m

e
 (

%
)

0

1

2

3

4

5

6

WM

CL

 

Figure 1: Particle size distribution of oil suspensions treated with high-shear mixing (dashed lines) and high-

pressure homogenization (solid lines). 

Table 1: Effect of different mechanical treatments on WM and cellulose-based oil dispersed on particle size 

distribution expressed as the main characteristic diameters. 

Component 
WM CL 

HSM HPH HSM HPH 

d(0.1) 6.4 ± 0.3 5.5 ± 0.2 11.6 ± 0.4 10.2 ± 0.5 

d(0.5) 101.3 ± 1.8 30.9 ± 1.9 43.2 ± 1.9 36.9 ± 2.7 

d(0.9) 461.3 ± 23.5 136.9 ± 7.3 168.0 ± 3.0 124.5 ± 7.5 

D[3,2] 15.1 ± 1.2 12.4 ± 0.7 17.5 ± 0.5 15.7 ± 1.0 

D[4,3] 177.8 ± 10.7 53.3 ± 2.7 76.3 ± 2.4 63.7 ± 2.5 

Table 1 reports the characteristic diameters of WM and CL oil suspensions, treated by HSM and HPH. As shown 

also in Figure 1, the mean particle size of WM was significantly reduced after HPH treatment: D[4,3] decreased 

by ~70%, D[3,2] by ~17%, and significant decreases were observed for both d(0.5) and d(0.9). In contrast, HPH 

treatment on cellulose-based oil suspension did not significantly change the size distribution with respect to 

HSM treatment: D[4,3] and d(0.5) decreased by ~15%, and D[3,2] by ~10%, whereas only the larger particles 

were effectively reduced in size, as shown by the decrease of d(0.9) of ~10% upon HPH processing. 

153



3.2 Capillary suspensions characterization 

The addition of water into the oil suspensions caused a drastic change in the system structure, due to the 

formation of capillary bridges and the development of a gel-like network, with higher consistency as the SR 

increased, as shown in Figure 2. At SR = 0.05 and 0.30, both WM and CL suspensions exhibited a weak gel-

like behavior, with no interconnected water-rich aggregates. With the increase in water concentration, an 

increase in aggregates was observed (the dark area in the microscopic images). Therefore, a compact and 

homogeneous 3D structure was developed for SR = 0.50, suggesting the formation of a capillary bridge network. 

 

Figure 2: Microscopic observations of capillary suspensions prepared from WM (first raw) and CL (second raw) 

oil suspensions adding water at different saturation ratios (SR = 0.05 in the first column, 0.30 in the second 

column, and 0.50 in the third column). 

The rheological behavior of WM and CL stabilized capillary suspensions is shown in Figure 3. In agreement with 

the results of the microscopic analysis in Figure 2, the formation of a 3D network structure was observed upon 

water addition, which became stronger with the increase of SR value, with a corresponding increase in apparent 

viscosity. Remarkably, for WM-stabilized capillary suspensions, higher apparent viscosity values than for CL-

stabilized suspensions were observed, at a low shear rate, for SR = 0.30 and SR = 0.50, whereas higher values 

were recorded for CL at SR = 0.05. Moreover, due to the gel-like structure of the suspensions, at all SRs, as the 

shear rate increased, the network broke down, and the viscosity decreased rapidly. 

Shear rate (1/s)

V
is

c
o

s
it
y 

(P
a

.s
)

S = 0.05

S = 0.30

S = 0.50

0.1 1 10 100 1000

0.01

0.1

1

10

100

1000

10000

Shear rate (1/s)

V
is

c
o

s
it
y 

(P
a

.s
)

0.1 1 10 100 1000

0.01

0.1

1

10

100

1000

10000

 

Figure 2: Flow curves of capillary suspensions stabilized by WM (left) and CL (right) particles as a function of 

the saturation ratio SR. 

154



Table 2 summarizes the results of the apparent yield stress values, derived from the flow curves, calculated as 

previously reported (Mustafa et al., 2018). The capillary suspensions exhibited an increase in apparent yield 

stress with increasing water concentration (SR) in the oil suspensions. This trend can be explained by the 

increase in network strength. However, when SR is increased above a critical value (SR > 0.5), a phase 

inversion was observed, with a switch from a continuous oil phase to a continuous water phase (data not 

reported). Moreover, using different types of solid particles (e.g. WM or CL) greatly affected the network 

formation in the capillary suspensions, and the deriving rheological properties: the apparent yield stress of CL-

based capillary suspension was significantly higher (one order of magnitude) than for WM at SR = 0.05, but it 

was significantly lower (one order of magnitude) at SR of 0.50. The differences in the rheological behavior of 

the different particles can be assumed to be related to the shape of the particles, affecting the capillary bridges 

formed in the suspensions (Maurath et al., 2016). 

Table 2: Apparent yield stress for WM and CL-in-oil suspensions at 30% and 5% w/w, respectively, as a 

function of saturation ratio SR. 

SR (-) 
Apparent yield stress (Pa) 

WM CL 

0.05 1.10 ± 0.32aA 11.70 ± 2.24aB 

0.30 33.26 ± 7.01bA 25.14 ± 4.98bA 

0.50 322.23 ± 22.30cB 31.20 ± 0.34cA 

Different lowercase letters denote statistically significant (p < 0.05) differences in the same column. 

Different capital letters denote statistically significant (p < 0.05) differences in the same raw. 

3.3 Capillary suspensions’ oxidative stability 

The 15-day period of storage in the accelerated Schaal oven test conditions (65 °C) correspond to 15 equivalent 

months at room temperature (Lužaić et al., 2022). A remarkable increase in peroxide values, which are the 

indicator of primary lipid oxidation, was noticed both in pure sunflower oil and CL-based capillary suspensions 

for all tested SRs, with higher values in CL-based capillary suspension because of the presence of water, which 

enhanced the oxidation process. In contrast, WM-based capillary suspensions ensured always the lowest PV 

in the oil: the phenolic bioactive compounds with high antioxidant activity, contained in the WM particles, but not 

present in the CL capillary suspensions, enabled to significantly slow down the oil oxidation phenomena during 

the entire period of storage, despite the presence of water, demonstrating that the proposed approach not only 

does not suffer the presence of water, but significantly improves the oxidation stability with respect to the pure 

oil. 

Table 3: Peroxide value (PV) of capillary suspensions after a 15-day storage period at 65 °C. 

SR (-) 
PV (mEq kg−1

oil) 

WM CL 

0.00* 62.41 

0.05 16.01 104.55 

0.50 13.02 99.52 

* pure sunflower oil, without suspended particles 

4. Conclusions 

The present work showed the feasibility of a novel approach proposed for oil structuring, based on direct 

processing via high-pressure homogenization of oil suspension of fibrous particles, such as wheat middlings, 

and subsequent addition of water to form a 3D network stabilized by interparticle capillary forces. 

This approach offers many advantages in comparison with more conventional approaches based on 

oleogelation, namely: (1) a facile and versatile process, based only on physical processing, without the addition 

of undesired ingredients to the formulation (e.g. gelators), (2) the replacement of part of the oil with the fibrous 

material and water, hence reducing the caloric intake, while ensuring the desired mouthfeel, as the oil will still 

constitute the continuous phase, (3) the delivery of bioactive molecules, associated with the fibours particles, 

when obtained from agri-food residues or other biomass, still rich in valuable compounds, with an antioxidant 

effect on the oil and expected health-beneficial effects also on consumers, and (4) the high transferability of the 

proposed approach, enabling to tune color, flavor, and taste of the structured oil, depending on the desired 

application. Currently, the main limitation of the proposed approach is the reduced temperature dependence of 

155



the rheological behavior, which could affect the organoleptic properties in some applications (e.g. when the solid 

fat are expected to melt in the mouth), and which needs to be addressed by future studies. 

Acknowledgments 

This research was funded by the Italian Ministry of University (MUR) call PRIN 2017 with the project 

2017LEPH3M “PANACEA: A technology PlAtform for the sustainable recovery and advanced use of 

NAnostructured CEllulose from Agro-food residues”. 

References 

Blake AI., Toro-Vazquez JF., Hwang HS., 2018, Wax oleogels (eds. Garti N. and G. Alejandro M.). San Diego, 

CA: AOCS Press. 

Bossler F., Koos E., 2016, Structure of Particle Networks in Capillary Suspensions with Wetting and Nonwetting 

Fluids, Langmuir, 32, 1489-1501 

Bossler F., Weyrauch L., Schmidt R., Koos E., 2017, Influence of mixing conditions on the rheological properties, 

and structure of capillary suspensions, Colloids and Surfaces A: Physicochemical and Engineering Aspects, 

518, 85-97. 

Jurić S., Ferrari G., Velikov KP., Donsì F., 2019, High-pressure homogenization treatment to recover bioactive 

compounds from tomato peels, Journal of Food Engineering, 262, 170-180. 

Koos E., 2014, Capillary suspensions: Particle networks formed through the capillary force, Current Opinion in 

Colloid and Interface Science, 19, 575-584. 

Le Révérend BJD., Norton IT., Cox PW., Spyropoulos F., 2010, Colloidal aspects of eating, Current Opinion in 

Colloid and Interface Science, 15, 84-89. 

Lužaić TZ., Grahovac NL., Hladni NT., Romanić RS., 2021, Evaluation of oxidative stability of new cold-pressed 

sunflower oils during accelerated thermal stability tests, Food Science and Technology, 42, 1-8 

Maurath J., Bitsch B., Schwegler Y., Willenbacher N., 2016, Influence of particle shape on the rheological 

behavior of three-phase non-Brownian suspensions, Colloids and Surfaces A: Physicochemical and 

Engineering Aspects, 497, 316-326 

Mustafa W., Pataro G., Ferrari G., Donsì F., 2018, Novel approaches to oil structuring via the addition of high-

pressure homogenized agri-food residues and water forming capillary bridges, Journal of Food Engineering, 

236, 9-18. 

Pirozzi, A., Capuano, R., Avolio, R., Gentile, G., Ferrari, G., Donsì, F., 2021, O/W pickering emulsions stabilized 

with cellulose nanofibrils produced through different mechanical treatments, Foods, 10, 1886. 

Zhao W., Wei Z., Xue C., 2021, Recent advances on food-grade oleogels: Fabrication, application and research 

trends, Critical Reviews in Food Science and Nutrition, 62 ,7659-7676. 

156


	29pirozzi.pdf
	Structuring Vegetable Oils Through the Formation of Capillary Suspensions: Comparison of Wheat Middlings and Pure Cellulose Processed by High-Pressure Homogenization