Abstract
Gelatine-gum Arabic coacervated microcapsules were prepared using complex coacervation followed by spray drying, with L-carvone, limonene, or hexyl salicylate encapsulated as a model flavour or fragrance oil. The influences of different types of core oil or further treatments including maltodextrin coating or glutaraldehyde crosslinking on size, morphology, payload, encapsulation efficiency and mechanical strength of the resulting microcapsules were investigated. Spherical microcapsules with polynuclear core-shell structure were fabricated. The encapsulation efficiency ranged from 5.0 ± 0.4% to 89.0 ± 1.2% for different core oils and 21.1 ± 0.1% to 89.0 ± 1.2% after different treatments. The highest encapsulation efficiency of 89.0% ± 1.2% was obtained for encapsulating hexyl salicylate with a hardening process by glutaraldehyde crosslinking. Micromanipulation tests revealed that gelatine-gum Arabic microcapsules exhibited typical rupture behaviours under compression. The rupture force increased with microcapsule diameter, the nominal rupture stress and toughness decreased with the increase of microcapsule size, and the nominal rupture tension was independent of diameter. The apparent Young’s modulus of whole microcapsules was determined by the Hertz model fitted to the experimental force versus displacement data. It has been found that different types of core oils significantly impacted their encapsulation efficiency, the structures of the formed microcapsules, and consequently mechanical properties.
Introduction
L-carvone, limonene, and hexyl salicylates are widely used as flavours and fragrances. However, these aroma molecules are highly volatile, as well as thermally sensitive and/or chemically reactive (Tekin et al., 2013) and challenging to be incorporated directly into products as they are either prone to evaporating during the manufacturing and storage process or reacting with other formulated ingredients. Moreover, these oils may undergo degradation or autoxidation under harsh conditions, such as exposure to light, heat, and oxygen as well as the variations in pH and humidity (Sousa et al., 2022). Therefore, microencapsulation is commonly employed to inhibit their evaporation and/or chemical degradation in formulated products for the industrial application of these oils.
Partanen et al. (2002) investigated the encapsulation of carvone by molecular inclusion, and found that the complexes of carvone in β-cyclodextrin exhibited good performance against core evaporation and showed superior storage property at higher relative humidities than spray dried maltodextrin-based microcapsules. However, molecular inclusion was not widely used in industry due to the regulatory restrictions in several countries (Desai & Jin Park, 2005). Besides, the high cost of cyclodextrins and low core loading capacities due to the stoichiometry of cyclodextrin molecules are inevitable problems limiting the application of this process (Reineccius, 1989). Koo et al. (2014) encapsulated peppermint oil using a coaxial electrospray system based on the gelation between calcium ions and alginate pectin mixture. Core-shell structured microcapsules were fabricated, with core oil encapsulation efficiency (EE) varying from 27% to 85%, which was attributed to the different ratios of alginate to pectin used. However, such process has an inevitable drawback due to its low throughput to fabricate the microcapsules. Baiocco et al. (2021b) encapsulated L-carvone via complex coacervation using fungal chitosan and gum Arabic (GA) as wall (shell) materials, followed by spray drying. The physical, morphological, structural, mechanical, and barrier properties of the fabricated microcapsules were systematically investigated for personal care and cosmetic products. However, L-carvone EE and the payload of the fabricated microcapsules were only 29 ± 4% and 19 ± 3%, respectively. Comparatively, Pakzad et al. (2013b) fabricated peppermint oil microcapsules with EE up to 82% using gelatine (GE) and GA coacervates as shell material. Dong et al. (2007) investigated the effect of different peppermint oil to GE and GA ratios on core oil loading of the microcapsules, and obtained the highest payload (90%) using the ratio of 6:1.
The encapsulation of limonene was studied using β-cyclodextrins via molecular inclusion (Astray et al., 2010; Dos Passos Menezes et al., 2017); however, only approximately 5% payload of limonene was detected by thermal analysis. Mallardo et al. (2016) attempted to improve the thermal stability of cyclodextrin–limonene inclusion complexes by incorporation of these molecules into biodegradable poly(butylene succinate) film, however, the payload was still relatively low (5%) due to the stoichiometry cavity of cyclodextrins. Spray drying was also widely used for the encapsulation of limonene with various wall materials including GA, modified starches, maltodextrin, or whey and soy proteins (Charve & Reineccius, 2009; Fisk et al., 2013; Ordoñez & Herrera, 2014; Paramita et al., 2012; Verdalet-Guzmán et al., 2013), however, the high temperatures involved in this process can lead to the volatilisation and/or oxidation of limonene molecules (Ibáñez et al., 2020). Yilmaz et al. (2023) encapsulated limonene using polyurethane-urea shell via interfacial polymerisation. Micro-sized capsules with spherical smooth morphology were fabricated, with limonene EE varying from 73% to 96%, based on different guanidine carbonate ratios in the polyurethane-urea shell. However, in this method, polyurethane was produced by reacting polyethylene glycol with isocyanate, which was reported to cause health problems including asthma, skin sensitisation, or cancer (U.S. Department of Labor, 2023). Baiocco and Zhang (2022) replaced polyurethane–urea by using fungal chitosan and GA as wall materials, and fabricated limonene laden microcapsules via complex coacervation and subsequent vacuum filtration drying. However, the core oil EE and payload of the fabricated microcapsules were only 44 ± 3% and 29 ± 2%, respectively.
Tasker et al. (2016) encapsulated hexyl salicylate in poly(methyl methacrylate) with hexadecyltrimethylammonium bromide as an emulsifier via solvent evaporation, while their research focussed only on the morphology of microcapsules without characterising oil EE or the mechanical properties of microcapsules. The microencapsulation of hexyl salicylate, on the other hand, was extensively investigated using formaldehyde-based material as a shell via in situ polymerisation (He et al., 2019; Long et al., 2009; Luo et al., 2022; Mercadé-Prieto et al., 2012). The properties of the microcapsules such as morphology, size and size distribution, shell thickness and permeability, mechanical strength, and EE were comprehensively studied. However, these materials used for hexyl salicylate encapsulation are synthetic polymer based, which can be classified as micro-plastic materials. Instead, Baiocco et al. (2021a) encapsulated hexyl salicylate using biopolymers based on the complex coacervation between chitosan and GA. Although the coacervated microcapsules were reported to have similar mean size and mechanical properties to melamine formaldehyde microcapsules, the EE (47%) was significantly lower than the formaldehyde-based microcapsules prepared by Luo et al. (2022) and Long et al. (2009), which had an EE up to 89% and 75%, respectively.
As can be seen from the above review, the outcomes of encapsulation strongly depended on the specific core oil, shell material and the process used to produce microcapsules. GE and GA as shell material of microcapsules to encapsulate flavours and fragrances by complex coacervation has received the most attention. However, there has been a lack of a systematic study on the effect of core oil on various properties of microcapsules prepared using the same shell material and process.
Moreover, understanding the mechanical properties of microcapsules is crucial to prevent their breakage in processing to produce finished projects, and to control the release of their active ingredient at the end-use applications. Yu et al. (2021) characterised the mechanical properties of GE/GA microcapsules when encapsulating functional fish oil. The Young’s modulus of whole microcapsules was determined to be 313 ± 42 MPa by the Hertz model. The rupture force increased with microcapsule diameter, with an average value of 0.58 ± 0.10 mN for the mean diameter of 10.8 ± 0.9 μm. The corresponding nominal rupture stress was 6.1 ± 0.6 MPa and EE was 89 ± 1%, which are higher than chitosan-gum Arabic coacervated microcapsules fabricated for L-carvone encapsulation by Baiocco et al. (2021b), which had a nominal rupture stress of 2.1 ± 0.3 MPa and EE of 29 ± 4%.
So far, information about the mechanical properties of microcapsules made of the same shell but different core coils is very limited.
In this study, each of three different oils was encapsulated initially using coacervation of GE and GA, followed by spraying coating of maltodextrin or crosslinking with glutaraldehyde, and the EE, size, morphology, structure, and mechanical properties were investigated with a focus to understand the effect of core oil on the EE, structure, and the mechanical properties of the resulting microcapsules.
Materials and methods
Materials
L-carvone (LC, 99%, w/w), S-limonene (LM, 95%, w/w), hexyl salicylate (HS, 98%, w/w), and glutaraldehyde (50%, w/w) were all purchased from Macklin, China. GA was purchased from Solarbio Life Sciences, China. GE (type B, ∼220 g bloom, average Mw = 500 kDa) was obtained from Shanghai Yuanye Biotechnology Co. Ltd., Shanghai, China, and maltodextrin (MD, DE16.5–19.5) was purchased from Sigma-Aldrich, UK. Reagents including absolute ethanol and 1-propanol were purchased from Sigma-Aldrich, UK, were of analytical grade and used without further purification. All the solutions were prepared using deionised water (18.2 MΩ cm at 25 °C).
Preparation of microcapsules
The GE-GA coacervated microcapsules were prepared based on the method (Yu et al., 2017; Yu et al., 2021) with some modifications. The aqueous phase was prepared by dissolving 2.5 g GE and 2.5 g GA in deionised water (200 ml) at 60 °C with magnetic stirring for 1 hr. The resulting solution was continuously stirred for 12 hr at 25 ± 2 °C. About 2.5 g of model oil (L-carvone, limonene, or hexyl salicylate) was added to the GE-GA mixture solution and homogenised by a high shear mixer (Model L4RT, Silverson, UK) at 3000 rpm for 10 min to prepare an oil-in-water emulsion. During the homogenisation, the beaker was covered with aluminium foil. After emulsification, the obtained emulsion was transferred into a beaker equipped with four standard baffles and kept under mechanical agitation at 400 rpm. In the meantime, the pH of emulsion was adjusted to 4.0 by adding 10% aqueous acetic acid solution, which induced the complex coacervation occurring at the surface of oil droplets by the formation of coacervates between GE and GA due to their opposite electrostatic charges. The oil loaded GE-GA microcapsules can be obtained under continuous agitation at 400 rpm for further 4 hr.
Microcapsules prepared via direct spray drying
The obtained HS-loaded microcapsule suspension was fed into a mini spray dryer YM-8000B (YUMING, Shanghai, China) to produce dry microcapsules, denoted by GG-HS. The suspension was atomised into a cylindrical chamber through a stainless-steel nozzle (2.8 mm in diameter). The spray drying system was operated with an inlet air temperature at 180 ± 2 °C and drying air flow rate at 35 |${\mathrm{m}}^3$|/hr. The emulsion was kept under stirring with a feeding flow rate of 2 ml/min into the drying chamber, and the resulting outlet air temperature was recorded to be 80 ± 5 °C. Free-flowing powders of microcapsules were collected from the bottom container and stored in glass vials for further analysis.
Microcapsules prepared with additional coating
About 1.5 g MD was dissolved in 100 ml deionised water at room temperature with magnetic stirring to prepare the coating solution. About 50 ml obtained HS-loaded microcapsule suspension was added to the coating solution under agitation. Then the mixed suspension was fed into the spray dryer to produce double shell microcapsules following the procedure in section Microcapsules prepared via direct spray drying, denoted by GG-HS-MD.
Microcapsules prepared with a crosslinker
After complex coacervation, the obtained oil-loaded microcapsule suspensions were added by 0.5 ml 50% aqueous glutaraldehyde as a crosslinker to harden GE-GA coacervated gelling shell, and the microcapsule suspension was agitated for 12 hr followed by subsequent spray drying described in section Microcapsules prepared via direct spray drying. Samples are donated by GG-LC-GT, GG-LM-GT, and GG-HS-GT for LC, LM, and HS oil, respectively. As control sample, the crosslinked GE-GA microcapsules containing no oil were also prepared following the same procedures. The detailed ingredients are listed in Table A-1, See online supplementary material. Each sample was prepared in triplicate.
Particle size analysis
Before filtration and drying, the mean size and size distribution of the fabricated microcapsules in suspension were measured using a light scattering instrument (Mastersizer 2000, Malvern Instruments Ltd, UK) with the Mastersizer 2000 software. The measurements were aligned by a reference refractive index of 1.505 for HS, 1.495 for LC, and 1.473 for LM, respectively (Baiocco et al., 2021a, 2021b; Clará et al., 2009). The refractive index of water was set at 1.330 and the refractive index of GE-GA coacervates gel was tested to be 1.59 by a J357 automatic refractometer, and the data are in coincident with literature (Rousi et al., 2019; Yang et al., 2012). The mean diameter and size distribution of microcapsules were measured in triplicate with a 2-min interval between each measurement. The values of Sauter mean diameter D[3,2] and SPAN were recorded, and they are defined as follows (Pan et al., 2013):
where di represents the diameter of an individual particle and N is the total number of particles measured, D90%, D50%, and D10% represent the diameter of particles under which the accumulative volume fraction is 90%, 50%, and 10%, respectively.
Morphological characterisation of GG-GA coacervated microcapsules
The morphology of microcapsules was firstly observed by a Leica DMRBE optical microscope (Leica microsystems, USA) with A Moticam Pro 252B camera fitted on and the images of microspheres were captured by Motic Images Advanced 3.2 software. Further morphological characterisations were performed by a scanning electron microscope (SEM, Hitachi TM3030 Tabletop). Five different spray dried microcapsules were tapped on the adhesive carbon tabs that were attached on the steel specimen stubs respectively. Samples were coated with platinum by using a sputter coater (Quorum SC7620) in order to make the targets electrically conductive so that high resolution images can be obtained. All the measurements were operated at an acceleration voltage of 20 kV.
Payloads and encapsulation efficiency
Sample preparation and calculation
Each batch of 20 mg spray dried oil-loaded microcapsules was firstly placed into a screw capped glass vial, and then dispersed into 20 ml of 36% aqueous 1-propanol used as the solvent for HS oil, whilst pure ethanol was used to extract LC or LM. The vials were ultrasonicated in a water bath (VWR Ultrasonicator, USC100TH, UK) for 60 min to ensure a complete extraction of each core oil. The insoluble coacervates shell residues were removed by centrifugation (Hermle Labortechnik Z-180, Labnet, Germany, EU) at 10,000 rpm (12,298 g force) for 30 min to produce a clear supernatant containing oil. The payload and EE were then calculated by the following equations:
The experiments for each oil were conducted in triplicate.
Empty GE-GA microcapsules containing no oil were also measured following the same procedures, to eliminate the absorbance influence from coacervates in solvents. And the corrected absorbance (λ) of each encapsulated oil was calculated by the following formula:
All the results showed that the shell coacervates did not dissolve in the solvent and did not consequently influence the UV absorption readings.
UV–Vis spectrophotometry to quantify HS
Based on the peak absorbance wavelength of HS in 36% (w/w) 1-propanol (306 nm), a linear standard calibration curve between UV absorbance readings (y) and oil concentrations (x) in 20 ml solvent was generated, |$y=21.286x+0.0314$| with a coefficient of determination |${R}^2=0.9994$| for HS oil. The obtained calibration equation was further used to determine the amount of encapsulated oil in microcapsules by the UV–Vis spectrophotometer (Cecil 1020, Cecil Instruments, UK) (Baiocco et al., 2021a). The resultant supernatant for HS samples was diluted in a quartz cuvette 10 times using 36% aqueous 1-propanol prior to measurement.
Gas-chromatography to quantify LC and LM
Since LC and LM are highly volatile and the absorption peak wavelength (218 nm) of LM positioned within the background noise area (200–300 nm) of Cecil 1020 UV–Vis spectrophotometer (data not shown), the quantitative analyses of LC and LM were performed by a gas chromatography method and ethanol was used as the solvent for LC and LM. Linear standard calibration curves for LC and LM were determined based on the rising areas (y) from detection peak and oil concentrations (x) in 20 ml ethanol: |$\mathrm{y}=19.075\mathrm{x}-7.2779$| with a coefficient of determination R2 = 0.9998 for LC and |$\mathrm{y}=15.821\mathrm{x}+2.7718$| with a coefficient of determination R2 = 0.9994 for LM, which were further used to determine the amount of encapsulated LC or LM in dry microcapsules measured by gas chromatography (GC) using an Agilent GC 8890 system (Agilent, USA) coupled with an Agilent 7693 automatic sample injector. The supernatants prepared as described in Section 2.5.1 for LC or LM were transferred into 2 ml glass vials (Agilent, USA) sealed with rubber caps (Thermo scientific, UK). The GC tests were performed with an Agilent HP-5 capillary column (30 m × 0.32 mm × 0.25 μm, Agilent, USA).
Instrumental parameters for LM analysis from the standard operating procedure of the manual were as follows: the temperature of the injector was 250 °C; N2 flow rate was 1.5 ml/min; column temperature program: initial temperature at 50 °C, increased to 180 °C at 10 °C/min (held for 2 min), and then increased at 5 °C/min to 200 °C (held for 1 min); the temperature of the detector was 250 °C. The separation time of LM was at 10.5 min based on the response of LM to the detector in the column.
Parameters for LC analysis were as follows: the temperature of the injector was 250 °C; N2 flow rate was 1.5 ml/min; column temperature program: initial temperature at 100 °C, increased to 230 °C at 10 °C/min (held for 5 min), and then increased at 10 °C/min to 240 °C (held for 1 min); the temperature of the detector was 250 °C. The separation time of LC was at 13.5 min based on the response of LC to the detector in the column.
Mechanical characterisation of microcapsules
Micromanipulation
The mechanical properties of GE-GA coacervated microcapsules were characterised by a micromanipulation technique as shown in Figure A-1, See online supplementary materia, which was firstly developed at the University of Birmingham UK (Zhang et al., 1991; Zhang, 1999; Hu et al., 2009). A single fine glass probe about 60 μm in diameter with a polished flat end surface was glued on a force transducer (GS0–10, Sensitivity 7.4225 mN V−1, LLC, Temecula, CA, USA), and the probe was positioned perpendicular to the glass slide, which was mounted on the stage of a micromanipulation rig. Spray dried microcapsules were spread or microcapsule suspension was dropped on the glass slide and observed from the side view camera, which was also used to measure the diameter of selected capsule and record the process of compression. Single coacervated microcapsules were compressed at 2 μm/s speed by the glass probe driven by the stepping motor. At least 30 microcapsules from each sample were randomly selected for compression. Compliance tests were performed in triplicate prior to testing each sample, and their mean values were used to calculate the actual displacement of the force probe (Zhang et al., 2022).
Figure 1 presents a typical compression process for an individual microcapsule before and after compression from the micromanipulation rig. The voltage signals generated by the force transducer due to compression were simultaneously recorded by the computer with a data acquisition card installed in it. The compression force could be calculated based on the sensitivity of force transducer and change in voltage recorded by the data acquisition card. From the curve of compression force versus displacement of the glass probe, the relationship between the force and the particle deformation could be obtained using a data analysis software package as reported by Zhang et al. (2022). Accordingly, the values of rupture force (|${F}_r$|) and displacement (|${\delta}_r$|) at rupture for the selected microcapsule can be determined, and the corresponding fractional deformation at rupture (|${\varepsilon}_r$|), nominal rupture stress (|${\sigma}_r$|), nominal rupture tension (|${T}_r$|), and toughness (|${T}_c$|) can be calculated based on the following equations (Zhang et al., 2022):
where subscript r represents the rupture point, R is the initial diameter of the tested microcapsule, |$\sigma$| is the nominal stress, |${T}_r$| is the nominal rupture tension, |$\varepsilon$| is the fractional deformation and |${T}_c$| is toughness that corresponds to the integration of the nominal stress over the fractional deformation up to rupture using equation (9).
Images of a GG-LM-GT microcapsule (13.2 um) before compression (A) and after compression (B). Scale bars represent 20 μm.
Hertz model
The Hertz model presents the relationship between the force for compressing a single linear elastic spherical particle to small displacement between two rigid flat planes (usually less than nominal strain (fractional deformation) 10%), and is described by Equation (10) (Yan et al., 2009; Zhang et al., 2009):
where F is the compression force, δ is the displacement, R is the particle diameter, υ is the Poisson's ratio, and E is the Young's modulus of the particle. According to the model, there should be a linear relationship between F and |${\delta}^{\frac{3}{2}}$|, and the value of Young's modulus can be determined by fitting the force-displacement data in elastic range (usually up to 10% of fractional deformation) to Equation (10) if Poisson's ratio υ is known. In this study, the Poisson’s ratios of all samples were assumed to be 0.5 since the GA-GE microcapsules were considered to be incompressible.
Statistical analysis
Three independent replicates for each experiment were performed. The difference among the results was expressed as mean value ±2 × standard error (St.Err).
Results and discussion
Morphology of microcapsules
The light microscopy images of different oil microcapsules prepared by coacervation are shown in Figure 2. It can be clearly seen that the suspended microcapsules containing different core oils were of spherical shape and in micro size. The oil droplets were encapsulated within a GA-GE coacervate shell, forming multinuclear structured microcapsules regardless of the type of core oils. Similar morphological characteristics were also reported previously on encapsulation of castor oil in GA-GE coacervated microcapsules (da Silva et al., 2015), Michelia alba D.C. extract aromas (Samakradhamrongthai et al., 2019), peppermint oil (Dong et al., 2011), paprika oleoresin (Alvim & Grosso, 2010), and ascorbic acid (Comunian et al., 2013). Dong et al., (2011) investigated the effect of core/wall weight ratio on the morphology of GA-GE coacervated microcapsules and found that higher core/wall ratios resulted in microcapsule structural transition from spherical to irregular, which is coincident with reports by Baiocco et al., (2021a, 2021b) and Baiocco and Zhang (2022) when they fabricated irregular eye-shaped microcapsules with a core/wall ratio as high as 12:1, for the encapsulation of L-carvone, limonene, and hexyl salicylate via complex coacervation using chitosan and GA.
Optical microscopy images of different microcapsules prior to spray drying (A) LC, (B) LM and (C) HS. Scale bars represent 20 μm.
SEM micrographs of spray dried microcapsules with different cores are presented in Figure 3. As can be seen in Figure 3A, C, E, microcapsules with different oils appear to have relatively spherical shapes and smooth surfaces, with the presence of wrinkles or concavities, which may be a consequence of the shrinkage of microcapsules due to the rapid evaporation of moisture during spray drying (Rocha et al., 2012) and vaccum drying during SEM imaging.
SEM images of (A) spray dried GG-LC-GT microcapsules, (B) partly incomplete GG-LC-GT microcapsules, (C) GG-LM-GT microcapsules, (D) partly incomplete GG-LM-GT microcapsules, (E) GG-HS-GT microcapsules, (F) partly incomplete GG-LM-GT microcapsules.
Similar morphological characteristics were also reported by other authors using the same wall materials for encapsulation of fish oil (Yu et al., 2021), and peppermint oil (Dong et al., 2011) via GA-GE complex coacervation. In addition, similar morphologies were also observed for spray dried microcapsules using different shell materials, when orange essential oil was encapsulated by whey protein isolate and GA as wall materials (Rojas-Moreno et al., 2018), β-carotene was encapsulated by modified tapioca starch (Loksuwan, 2007), and ascorbic acid was encapsulated by Capsul and maltodextrin (Finotelli & Rocha-Leão, 2005). It was further demonstrated by Sheu & Rosenberg (1998) that the formation of indentations on microcapsule surface can potentially be attributed to wall compostions and spray drying conditions. Besides, multiple minute coacervates with smooth surface were also observed at the surface of relatively larger microcapsules. Partly incomplete microcapsules with their unsealed shell can be seen in Figure 3B, D, F, suggesting the core-shell inner structure for different oil microcaspules.
Particle size and size distribution
The size and size distribution of HS oil droplets and HS microcapsules are illustrated in Figure 4. The surface mean size of HS oil droplets in emulsion is approximately 3.85 ± 0.01 μm whilst HS microcapsules have an average size of 10.71 ± 0.06 μm, further suggesting the successful entrapment of several core oil droplets within the shell materials.
Size and size distribution of HS emulsion droplets and HS microcapsules.
The mean size and size distribution of different oil microcapsules are displayed in Table 1. GG-LM-GT and GG-HS-GT microcapsules have similar average size of 11.0 ± 0.2 and 10.7 ± 0.1 μm, respectively, which are relatively greater than GG-LC-GT microcapsules (9.8 ± 0.1 μm). In addition, all samples exhibit narrow size distributions with a relatively small span value approximately at 0.7. Theoretically speaking, the size of microcapsules is largely determined by that of primary emulsion droplets and their possible coalescence, while the size of emulsion droplets can be related to the interfacial tension between the core oil and the aqueous phase, its viscosity and specific density (Padron and Calabrese, 2023). The large interfacial tension, high viscosity, or big difference in specific density between the oil and aqueous phase tends to lead bigger size of emulsion droplets, which combined with the mixing condition determined the final size and size distribution of the microcapsules.
Mean surface size and span of different microcapsules (mean ± 2 × St. Err).
| Microcapsules | GG-LC-GT | GG-LM-GT | GG-HS-GT |
|---|---|---|---|
| |${D}_{\left[3,2\right]}$| (μm) | 9.8 ± 0.2 | 11.0 ± 0.2 | 10.7 ± 0.1 |
| Span | 0.739 ± 0.001 | 0.728 ± 0.001 | 0.778 ± 0.001 |
| Microcapsules | GG-LC-GT | GG-LM-GT | GG-HS-GT |
|---|---|---|---|
| |${D}_{\left[3,2\right]}$| (μm) | 9.8 ± 0.2 | 11.0 ± 0.2 | 10.7 ± 0.1 |
| Span | 0.739 ± 0.001 | 0.728 ± 0.001 | 0.778 ± 0.001 |
Mean surface size and span of different microcapsules (mean ± 2 × St. Err).
| Microcapsules | GG-LC-GT | GG-LM-GT | GG-HS-GT |
|---|---|---|---|
| |${D}_{\left[3,2\right]}$| (μm) | 9.8 ± 0.2 | 11.0 ± 0.2 | 10.7 ± 0.1 |
| Span | 0.739 ± 0.001 | 0.728 ± 0.001 | 0.778 ± 0.001 |
| Microcapsules | GG-LC-GT | GG-LM-GT | GG-HS-GT |
|---|---|---|---|
| |${D}_{\left[3,2\right]}$| (μm) | 9.8 ± 0.2 | 11.0 ± 0.2 | 10.7 ± 0.1 |
| Span | 0.739 ± 0.001 | 0.728 ± 0.001 | 0.778 ± 0.001 |
Payload and encapsulation efficiency
The payload and EE of different microcapsules are summarised in Table 2. As can be seen, the payload and EE of GG-HS-GT were determined to be 29.7 ± 0.4% and 89.0 ± 1.2%, which are similar to the values (28.3 ± 0.7% and 89.8 ± 1.4%) of fish oil microcapsules prepared by Yu et al. (2021) following the same experimental formulations and procedures. Similar EE values were also reported for the encapsulation of other essential oils employing the coacervation between GA and GE by other authors (Khatibi et al., 2021; Napiórkowska et al., 2023; Pakzad et al., 2013a; Rungwasantisuk & Raibhu, 2020; Xiao et al., 2014). The obtained GG-HS-GT microcapsules present superiorly to the HS entrapped microcapsules (EE, 47 ± 11%) prepared by Baiocco et al. (2021a) using chitosan and GA as wall materials, indicating the superior performance of GE than chitosan in forming microcapsules with a higher EE (Baiocco et al., 2021b; Bruyninckx & Dusselier, 2019). Besides, it is worth noting that additional treatments to freshly coacervated microcapsules significantly improved the payload as well as EE. A similar improvement was also reported by Tello et al. (2016) that the EE of soy oil increased from 79.0% to 89.9% after microcapsules were crosslinked with glutaraldehyde. Liu and Jiang (2023) discussed the influence of crosslinkers on encapsulating β-ionone, and found microcapsules crosslinked with higher concentrations of glutaraldehyde exhibited higher EE of the core material. Dinarvand et al. (2005) explained that a higher amount of glutaraldehyde increased the density of polymer through the chemical coupling of aldehyde groups with the amino groups of the GE (Chang et al., 2006), and thus reduced macromolecular chains mobility and finally formed stable and rigid spheres, which was conducive for better entrapment of the core material. Maltodextrin as an outer layer of physical coating for coacervated microcapsules also enhanced the payload and EE of microcapsules. However, its payload or EE was significantly lower than that of GG-HS-GT, suggesting the higher permeability of maltodextrin coating than the chemical reticulation after glutaraldehyde crosslinking. In addition, it is notable that the payload of GG-HS-MD was expected to be lower than GG-HS microcapsules as 1.5 g maltodextrin was used and increased the value of total mass of particles in Equation (3). However, higher values of both payload and EE were obtained for GG-HS-MD, suggesting that the additional physical coating by maltodextrin significantly reduced the core oil loss during the spray drying process.
Payload and corresponding encapsulation efficiency of different microcapsules (mean ± 2 × St. Err).
| Microcapsules | Payload (%) | Encapsulation efficiency (%) |
|---|---|---|
| GG-HS | 7.0 ± 0.1 | 21.1 ± 0.1 |
| GG-HS-MD | 10.1 ± 0.3 | 54.0 ± 2.0 |
| GG-HS-GT | 29.7 ± 0.4 | 89.0 ± 1.2 |
| GG-LM-GT | 9.9 ± 0.9 | 30.0 ± 2.6 |
| GG-LC-GT | 1.7 ± 0.1 | 5.0 ± 0.4 |
| Microcapsules | Payload (%) | Encapsulation efficiency (%) |
|---|---|---|
| GG-HS | 7.0 ± 0.1 | 21.1 ± 0.1 |
| GG-HS-MD | 10.1 ± 0.3 | 54.0 ± 2.0 |
| GG-HS-GT | 29.7 ± 0.4 | 89.0 ± 1.2 |
| GG-LM-GT | 9.9 ± 0.9 | 30.0 ± 2.6 |
| GG-LC-GT | 1.7 ± 0.1 | 5.0 ± 0.4 |
Payload and corresponding encapsulation efficiency of different microcapsules (mean ± 2 × St. Err).
| Microcapsules | Payload (%) | Encapsulation efficiency (%) |
|---|---|---|
| GG-HS | 7.0 ± 0.1 | 21.1 ± 0.1 |
| GG-HS-MD | 10.1 ± 0.3 | 54.0 ± 2.0 |
| GG-HS-GT | 29.7 ± 0.4 | 89.0 ± 1.2 |
| GG-LM-GT | 9.9 ± 0.9 | 30.0 ± 2.6 |
| GG-LC-GT | 1.7 ± 0.1 | 5.0 ± 0.4 |
| Microcapsules | Payload (%) | Encapsulation efficiency (%) |
|---|---|---|
| GG-HS | 7.0 ± 0.1 | 21.1 ± 0.1 |
| GG-HS-MD | 10.1 ± 0.3 | 54.0 ± 2.0 |
| GG-HS-GT | 29.7 ± 0.4 | 89.0 ± 1.2 |
| GG-LM-GT | 9.9 ± 0.9 | 30.0 ± 2.6 |
| GG-LC-GT | 1.7 ± 0.1 | 5.0 ± 0.4 |
Significant variations of payload and EE were obtained for different oil microcapsules following the identical formulations of the wall materials and preparation procedures. The highest payload and EE are 29.7 ± 0.4% and 89.0 ± 1.2% for GG-HS-GT, while GG-LM-GT presents a lower payload and EE at 9.9 ± 0.9% and 30.0 ± 2.6%, respectively, followed by GG-LC-GT of 1.7 ± 0.1% payload and 5.0 ± 0.4% EE. Such differences can be ascribable to several factors. HS has a vapour pressure of 0.0005 mm Hg at 25 °C, significantly lower than that of LM (1.98 mm Hg at 25 °C) and that of LC (0.115 mm Hg at 25 °C) (PubChem, 2023a; 2023b; Vigon, 2020). It is notably that GG-LM-GT presents higher values of payload and EE than GG-LM-GT although limonene has higher vapour pressure than L-carvone, suggesting it is unlikely that the loss of core oil was solely due to volatile evaporation. The LogP value of hexyl salicylate is 4.87, compared to LogP value of limonene at 4.57 and LogP value of L-carvone at 2.71 (Luo et al., 2023; Nikfar & Behboudi, 2014; PubChem, 2023b), which indicates that the significant loss of LC occurred due to its migration to the aqueous phase during encapsulation and/or spray drying process. Besides, it is presumed that the payload and EE of different oil microcapsules are also associated with their inner structures and morphologies, which can be influenced by the interfacial tension balance between different phases (oil phase, GA-GE coacervates phase and aqueous phase) (Loxley & Vincent, 1998; Torza & Mason, 1970).
Mechanical properties of coacervated microcapsules
Rupture analysis
Figure 5 presents a typical force versus displacement curve from compression of a representative GG-LM-GT microcapsule using the micromanipulation technique. Point a is the initial point when the probe touched the microcapsule, b is the yield point, and a-b curve corresponds to the elastic behaviour of the microcapsule. The curve of b-c shows the plastic behaviour of microcapsules and point c is the rupture point of the microcapsule (Luo et al., 2022). At point c, there was a significant instant force reduction due to the sudden burst of the microcapsule. The force increased gradually when compressing the ruptured microcapsule residue and dramatically when the force probe approached the substrate holding the sample of microcapsules.
Typical force versus displacement data obtained from the compression of an individual GG-LM-GT (13.2 μm in diameter) microcapsule. The dotted line indicates the rupture of the microcapsule.
The rupture force, displacement at rupture, nominal rupture stress, toughness, and nominal rupture tension versus diameter of different microcapsules are presented in Figure 6. The average values of the rupture force, displacement at rupture, nominal rupture stress, toughness, and nominal rupture tension were calculated and the data are presented in Table A-2, See online supplementary material. Both the rupture force and the displacement at rupture increase with microcapsule diameter. In contrast, the nominal rupture stress and toughness decreased with the increase of microcapsule size. The nominal rupture tension was independent of diameter. These trends are consistent with the previous work on microcapsules prepared with the same shell materials (Yu et al., 2021), or different wall materials (Baiocco et al., 2021a; Luo et al., 2022).
Mechanical strength parameters of microcapsules versus diameter: rupture force (A, a), displacement at rupture (B, b), nominal rupture stress (C, c), nominal toughness (D, d) and nominal rupture tension (F, f). 
: GG-HS, 
: GG-HS-MD, 
: GG-HS-GT, 
: GG-LC-GT, and 
: GG-LM-GT. The dotted lines only indicate the trend or the mean value for nominal rupture tension.
The mechanical properties of microcapsules varied significantly when different core oils or encapsulation methods were used. For microcapsules prepared with HS oil, GG-HS-MD and GG-HS-GT had similar values of toughness, which were significantly higher than that of GG-HS, indicating that microcapsules prepared with a second coating by maltodextrin or wall hardened by glutaraldehyde crosslinking were mechanically stronger than those without additional treatment. It is worth noting that a microcapsule rupturing at a relatively smaller deformation is more likely due to its higher structural brittleness. On the contrary, if a microcapsule ruptures at a larger deformation, the phenomenon is associated with its greater structural flexibility and stretchability (Baiocco & Zhang, 2022). The average rupture tension of GG-HS-MD was 175 ± 33 μN/μm, which is marginally higher than that of GG-HS-GT (120 ± 16 μN/μm), whilst the average deformation at rupture of GG-HS-MD (12 ± 1%) was slightly lower than GG-HS-GT (15 ± 1%). Although both additional treatments made microcapsules mechanically stronger, the former provided a layer of dried physical coating to make the microcapsules more structurally brittle while the latter made the microcapsules more flexible to resist capsule rupture through gelatine–glutaraldehyde crosslinking. For microcapsules prepared with different core oils, although there was no significant difference in the mechanical strength parameters between GG-LM-GT and GG-HS-GT, their rupture force, nominal rupture stress, rupture tension, and toughness of GG-LC-GT were significantly higher than the corresponding LM and HS microcapsules, which suggests that the influence of core oils on the mechanical strength of microcapsules is not negligible. Similar results were also reported by other authors (Baiocco & Zhang, 2022; Luo et al., 2022). Luo et al. (2022) encapsulated 2-hydroxy-3-(octanoyloxy) propyl decanoate, hexyl salicylate, lavender oil, and lily oil within melamine-glutaraldehyde-formaldehyde microcapsules via an in-situ polymerization method and found stronger microcapsules were obtained when encapsulating core oils with higher hydrophobicity. Baiocco & Zhang (2022) reported that LM microcapsules were mechanically weaker than the HS microcapsules after the wall was fabricated using GA and fungally fermented chitosan via complex coacervation and vacuum drying. They assumed the difference may be due to the nature of terpenic molecules (LM), which can cause a higher interfacial energy level and impair the intermolecular bonds within coacervates network, thereby affecting the overall robustness of the microcapsule.
The mechanical properties of microcapsules can be associated with their chemical compositions, geometries and inner structures, which resulted from the interfacial energy balance between the phases in the system (Loxley & Vincent, 1998; Torza & Mason, 1970). For instance, due to the more hydrophobic nature, HS has higher interfacial tension against aqueous phase than LC and LM, which favours the movement of coacervate phase to the water/oil interface to form the microcapsules with a core-shell structure. Comparatively, the lower value of interfacial tension between LC and aqueous phase results in less core-shell structured microcapsules, i.e., more solid-like microspheres, which correspondingly results in stronger mechanical strength. The results of payloads for each oil microcapsule possibly evidence this interpretation (see Table 2).
Apparent Young’s modulus determined by Hertz analysis
To further investigate the elastic behaviour of different microcapsules, the apparent Young’s modulus for each type of microcapsule was characterised by fitting the force versus displacement data corresponding to the nominal strain (factional deformation) up to 10% with the Hertz model. For each sample, more than 30 microcapsules were analysed and the Hertz model fitting to the force versus displacement data of a representative GG-HS-GT microcapsule is presented in Figure A-2, See online supplementary material, and the Young’s modulus for each type of microcapsules versus diameter is shown in Figure 7.
Apparent Young’s modulus of different microcapsules versus diameter (A) GG-HS, GG-HS-MD and GG-HS-GT and (B) GG-HS-GT, GG-LC-GT and GG-LM-GT. 
: GG-LC-GT, 
: GG-LM-GT, 
: GG-HS-MD, 
: GG-HS-GT, and 
: GG-HS.
Figure 7 presents the Young’s modulus versus particle diameter for the compressed microcapsules. As can be seen in Figure 7A, the Young’s modulus of GG-HS microcapsules was holistically lower than GG-HS-MD and GG-HS-GT, indicating the latter two types of microcapsules had higher stiffness. The mean values of Young’s modulus for each type of microparticle as well as the corresponding coefficient of determination (R2) are summarised in Table 3.
Mean values of apparent Young’s modulus for different microcapsules and their corresponding mean values of R2 from using the Hertz model (mean ± 2 × St.Err).
| Microcapsules | Young’s Modulus (MPa) | R2 |
|---|---|---|
| GG-HS | 163 ± 33 | 0.86 ± 0.04 |
| GG-HS-MD | 702 ± 125 | 0.81 ± 0.06 |
| GG-HS-GT | 668 ± 165 | 0.80 ± 0.04 |
| GG-LM-GT | 663 ± 79 | 0.80 ± 0.04 |
| GG-LC-GT | 891 ± 104 | 0.91 ± 0.04 |
| Microcapsules | Young’s Modulus (MPa) | R2 |
|---|---|---|
| GG-HS | 163 ± 33 | 0.86 ± 0.04 |
| GG-HS-MD | 702 ± 125 | 0.81 ± 0.06 |
| GG-HS-GT | 668 ± 165 | 0.80 ± 0.04 |
| GG-LM-GT | 663 ± 79 | 0.80 ± 0.04 |
| GG-LC-GT | 891 ± 104 | 0.91 ± 0.04 |
Mean values of apparent Young’s modulus for different microcapsules and their corresponding mean values of R2 from using the Hertz model (mean ± 2 × St.Err).
| Microcapsules | Young’s Modulus (MPa) | R2 |
|---|---|---|
| GG-HS | 163 ± 33 | 0.86 ± 0.04 |
| GG-HS-MD | 702 ± 125 | 0.81 ± 0.06 |
| GG-HS-GT | 668 ± 165 | 0.80 ± 0.04 |
| GG-LM-GT | 663 ± 79 | 0.80 ± 0.04 |
| GG-LC-GT | 891 ± 104 | 0.91 ± 0.04 |
| Microcapsules | Young’s Modulus (MPa) | R2 |
|---|---|---|
| GG-HS | 163 ± 33 | 0.86 ± 0.04 |
| GG-HS-MD | 702 ± 125 | 0.81 ± 0.06 |
| GG-HS-GT | 668 ± 165 | 0.80 ± 0.04 |
| GG-LM-GT | 663 ± 79 | 0.80 ± 0.04 |
| GG-LC-GT | 891 ± 104 | 0.91 ± 0.04 |
The mean R2 values obtained from all microcapsules are in a range of 0.80–0.91, which implies the fitting is good, but not perfect. The Hertz model can normally be applied to describe the relationship between the imposed compression force and the displacement of the spherical homogenous particles at small fractional deformation (up to 10%) (Baiocco et al., 2023; Wang et al., 2018). Herein, the calculated Young’s modulus in this work is in fact the apparent Young’s modulus of the whole microcapsules as they contained multiple oil droplets which are considered to have no elastic performance but are able to affect the force response under compression. The value of R2 of GG-LC-GT (0.91) is higher than those of other microcapsules, indicating that GG-LC-GT may be more homogenous and uniform in structure, which can be evidenced by the lower value of oil payload for GG-LC-GT microcapsules.
For HS microcapsules prepared by different methods, both GG-HS-MD and GG-HS-GT had significantly higher values of apparent Young’s modulus than GG-HS, indicating that the further treatments towards GE-GA coacervated microcapsules improved their elastic performance. Compared with GG-HS, the crosslink of GE/GA by glutaraldehyde resulted in a more compact shell structure for GG-HS-GT (Knaebel et al., 1997), potentially allowing higher payload of core oil. While for microcapsules prepared with identical shell materials, the mean apparent Young’s modulus of GG-LC-GT was significantly higher than GG-HS-GT and GG-LM-GT microcapsules. As aforementioned, core oils of different hydrophobicity may influence the diffusion of the coacervates from the bulk to oil/water interface or glutaraldehyde molecules and thus the crosslink reaction rate, leading to the gradient of local crosslink density and spatial distribution with the capsule shell (He et al., 2008), which eventually resulted in the variation in payload and mechanical properties even with identical shell chemical compositions.
Theoretically, Young’s modulus represents the intrinsic stiffness of an elastic material undergoing recoverable deformation, which is expected to be a constant value regardless of microcapsule size. However, as presented in Figure 7, it seems that the apparent Young’s modulus of microcapsules is strongly size dependant and larger microcapsules have less Young’s modulus values. As discussed above, the calculated apparent Young’s modulus is the outcome of liquid core combined with the shell material and cannot fully represent the intrinsic elastic modulus value of the shell material. For microcapsules with a simple core-shell, it is possible to determine the intrinsic Young’s modulus and the ratio of shell thickness to radius of single microcapsules by fitting experimental force versus displacement data to finite element simulation results (Mercadé-Prieto et al., 2011). However, the microcapsules formed in this study had multicores and more FEA work is required to handle microcapsules with such more complex structure.
Conclusion
In this study, different flavour/fragrance oils were encapsulated by the complex coacervation using GE and GA as wall materials. The influence of different core oils on the EE, payload as well as the physical, structural and mechanical properties of microcapsules were studied. It was observed that micro-size spherical microcapsules had polynuclear core shell structure. The highest EE of 89.0 ± 1.2% was obtained from GG-HS-GT, and evident variations on EE were calculated for different core oils. The EE of different core oil can be largely influenced by their LogP values and vapour pressures. Further treatments to coacervated microcapsules by spray coating with maltodextrin or glutaraldehyde crosslinking significantly improved EE. Mechanical characterisations of microcapsules were undertaken using micromanipulation measurements of their rupture strength parameters and determination of the apparent Young’s modulus of the whole microcapsule by the Hertz model. Additional processing by either the maltodextrin coating or glutaraldehyde crosslinking significantly enhanced the mechanical strength and stiffness of microcapsules. Different core oils affected the structure of the formed microcapsules, which consequently impacted their mechanical properties, including rupture strength parameters and apparent Young’s modulus. Future works include the investigation of the influence of core oil polarity and their interfacial tensions with aqueous phase and coacervate phase on the microcapsule morphology and corresponding EE.
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Author contributions
Qun Huang (Conceptualization, Methodology, Validation, Investigation, Formal analysis, Data curation, Visualization, Writing—original draft), Tom Mills (Supervision, Writing—review & editing), Zhibing Zhang (Conceptualization, Formal analysis, Resources, Visualization, Supervision, Writing—review & editing).
Funding
None declared.
Conflicts of interest
The authors declare no conflict of interest.
Acknowledgements
The authors would like to express their gratitude for the technical support provided by the staff of the School of Chemical Engineering, University of Birmingham, West Midlands, B15 2TT, UK.
