论文信息
论文解读
AI 辅助整理- 方法
- 结构化检索Scopus等三大数据库3105条,保留298篇做叙述性证据综合。
- 发现
- 提取物组成决定载体选择与包埋效果,高包埋率或小粒径不等于技术有效。
- 机会点
- 残渣源蛋白、果胶等作载体及安全、放大与经济技术环境评价仍属研究空白。
论文摘要
果蔬加工副产物是酚类、色素、精油及其他生物活性化合物的重要来源,但其作为食品配料的应用受到组成变异性、稳定性差以及许多化合物生物可及性低的限制。纳米包埋可解决其中部分局限,但回收与包埋通常被作为相互独立的操作进行研究。本综述将二者视为一个相互关联的过程,探讨副产物特性与回收条件如何决定回收物的组成,进而影响载体选择及其在食品中的性能。对Scopus、Web of Science和PubMed进行结构化检索,共获得3,105条记录,其中298条被保留用于叙述性证据综合。文献表明,品种、组织、成熟度和副产物稳定化处理会在提取前影响生物活性谱,而溶剂选择、纯化和预处理进一步决定分子形态、共提取物和残留溶剂。这些因素可改变包埋效率、载量、稳定性和胃肠道行为,并有助于解释为何使用看似相似的提取物所获得的结果并不总是可迁移;已报道的包埋效率约为14%至95%以上,但标志物和方法的差异使其无法直接排序。来自食品应用的证据进一步表明,载体性能取决于食品基质的组成和加工条件,仅凭高包埋效率或小粒径并不能确立技术优势。副产物来源的蛋白质、果胶、纤维素、木质素及全副产物材料也可用作载体。关于安全性、放大生产、技术经济可行性和环境性能的证据仍然有限,且大多数研究仍停留在实验室规模。基于上述发现,本文提出一个从回收到应用的框架,其中载体选择始于回收提取物的特性和目标食品的需求,而非包埋技术本身。
全文
Frontiers in Sustainable Food SystemsFrontiers in Sustainable Food Systems 01 frontiersin.org
TYPE Review
PUBLISHED 16 November 2026
DOI 10.3389/fsufs.2026.2008486
OPEN ACCESS
EDITED BY
Faizan Ahmad,
Aligarh Muslim University, India
REVIEWED BY
Miguel Ángel De León Zapata,
Autonomous University of Coahuila,
Mexico
Sujeeta Yadav,
Sanskaram University, India
CORRESPONDENCE
Poliana Mendes de Souza
poliana.souza@um6p.ma
RECEIVED 23 September 2026
REVISED 21 October 2026
ACCEPTED 23 October 2026
PUBLISHED 16 November 2026
CITATION
Babatunde J and de Souza PM (2026)
Nanoencapsulation of bioactive
compounds from fruit and vegetable
processing residues: linking recovery,
carrier selection, and food applications.
Front. Sustain. Food Syst. 10:2008486.
doi: 10.3389/fsufs.2026.2008486
COPYRIGHT
© 2026 Babatunde and de Souza. This is
an open-access article distributed under
the terms of the Creative Commons
Attribution License (CC BY). The use,
distribution or reproduction in other
forums is permitted, provided the
original author(s) and the copyright
owner(s) are credited and that the
original publication in this journal is
cited, in accordance with accepted
academic practice. No use, distribution
or reproduction is permitted which does
not comply with these terms.
Nanoencapsulation of bioactive
compounds from fruit and
vegetable processing residues:
linking recovery, carrier selection,
and food applications
Jubril Babatunde 1,2 and Poliana Mendes de Souza 1
1African Sustainable Agriculture Research Institute (ASARI), College of Agriculture and Environmental
Sciences (CAES), University Mohammed VI Polytechnic (UM6P), Laayoune, Morocco, 2Department of
Food Science and Technology, Ibrahim Badamasi Babangida University, Lapai, Nigeria
Fruit and vegetable processing residues are important sources of phenolics,
pigments, essential oils and other bioactive compounds, but their use as food
ingredients is limited by compositional variability, poor stability and, for many
compounds, low bioaccessibility. Nanoencapsulation can address some of these
limitations, although recovery and encapsulation are commonly investigated as
separate operations. This review examines them as a connected process, asking
how residue characteristics and recovery conditions determine the composition
of the recovered material and, in turn, influence carrier selection and perfor -
mance in foods. A structured search of Scopus, Web of Science and PubMed
identified 3,105 records, of which 298 were retained for narrative evidence syn-
thesis. The literature shows that cultivar, tissue, maturity and residue stabilisation
affect the bioactive profile before extraction, while solvent choice, purification
and pretreatment further determine molecular form, co-extractives and residual
solvent. These factors can alter encapsulation efficiency, loading, stability and
gastrointestinal behaviour and help explain why results obtained with apparently
similar extracts are not always transferable; reported encapsulation efficiencies
range from about 14% to above 95%, but differences in markers and methods
prevent direct ranking. Evidence from food applications further shows that carrier
performance depends on the composition and processing conditions of the food
matrix, and that high encapsulation efficiency or small particle size alone does
not establish technological benefit. Residue-derived proteins, pectins, cellulose,
lignin and whole-residue materials can also serve as carriers. Evidence for safety,
scale-up, techno-economic feasibility and environmental performance remains
limited, and most studies remain at laboratory scale. Based on these findings, a
recovery-to-application framework is proposed in which carrier selection starts
with the characteristics of the recovered extract and the requirements of the
target food rather than with the encapsulation technology itself.
KEYWORDS
bioaccessibility, bioactive compounds, circular economy, food applications, food
processing residues, green extraction, nanocarriers, nanoencapsulation
Babatunde and de Souza 10.3389/fsufs.2026.2008486
Frontiers in Sustainable Food Systems 02 frontiersin.org
1 Introduction
Fruit and vegetable processing leaves behind peels, pomaces,
seeds, husks and leaves that are richer in several classes of bioac -
tive compounds than the edible fractions sold as food. The obsta -
cle is not availability but performance: many of these compounds
are poorly soluble, unstable during processing and storage, and
poorly bioavailable ( Liu et al., 2024 ; Saini et al., 2019 ), and
reviews of waste valorisation add heterogeneous composition,
limited scalability and high cost ( Ingale et al., 2025 ; Y adav et
al., 2024 ).
Encapsulation protects labile compounds and can control their
release, and nanoscale systems add interfacial area that improves dis-
persibility, absorption and controlled release relative to microscale
carriers (Chellamboli et al., 2025 ). Small droplets also give higher
stability to aggregation and creaming and higher bioavailability of
lipophilic components, although the same features raise questions
about biological fate after ingestion (McClements and Rao, 2011). For
this review, ‘nanoscale’ is used operationally in the broader food-
colloid sense for carriers or dispersed structures with a reported mean
dimension below 1 μm. Systems above this limit are retained only
when they clarify terminology, boundary conditions or food
performance.
Most reviews focus on one compound, residue or delivery
system, while primary studies usually optimise extraction and
encapsulation separately ( Lv et al., 2026 ; Andishmand et al., 2023 ;
Saini et al., 2019 ; Sheikhzadeh et al., 2026 ). This approach treats
the extract as fixed even though residue type, cultivar, stabilisa -
tion and recovery method change its composition. As a result,
findings may not transfer between extracts, and carrier choice is
rarely linked to extract properties or the target food. Recent
reviews report high encapsulation efficiency (EE) and wide varia -
tion in bioaccessibility, but still present extraction and encapsula -
tion as separate outcomes ( Ligarda-Samanez et al., 2025 ;
Y ata-Franco et al., 2026 ; Marcillo-Parra et al., 2021 ). Only a few
studies treat delivery-system selection as a design problem involv -
ing the cargo, food matrix and gastrointestinal tract ( Sarkar et al.,
2020a; Assadpour, 2018 ). This review begins with the extract. It
examines how residue variability and recovery chemistry shape
carrier requirements, then follows those effects through carrier
selection, food application and digestion.
This review therefore treats residue characteristics and stabi -
lisation, recovery, carrier selection, food application, digestion
and safety as one linked design chain ( Figure 1 ). Its central con -
tribution is a carrier-selection framework that begins with the
recovered extract and the target food matrix (Section 4.3). A cir -
cularity lens runs through the framework because residues can
supply not only the bioactive cargo but also proteins, pectins,
cellulose, lignin or whole-residue materials for the carrier. This
dual use may reduce dependence on virgin materials, but it also
introduces further compositional and safety variability that must
be controlled.
The literature was assembled from Scopus, Web of Science and
PubMed. The core search combined four concept blocks: residue
source, bioactive cargo, nanoscale delivery system and the exclu -
sion of studies on the green synthesis of metal nanoparticles, and
was limited to English-language articles and reviews published
from 2015 onwards. Additional section-level searches on recovery,
carrier selection and food-matrix application were screened sepa -
rately against their own section questions and used as supporting
evidence.
The searches returned 3,105 records: 1,239 from Scopus, 1,050
from Web of Science and 816 from PubMed. After 663 duplicates
were removed using DOI, PubMed identifier and normalised title
matching, 2,442 unique records were screened on title and abstract
in two stages: a rule-based pre-screen against predefined inclusion
and exclusion criteria (residue source, bioactive cargo, nanoscale
delivery system, food relevance), followed by manual checking.
Overall, 1,941 records were excluded, most commonly because
they did not involve a nanoscale delivery system ( n = 906) or
focused on the green synthesis of metal nanoparticles using resi -
due extracts ( n = 300). Of the remaining 501 records, 298 (273
primary studies and 25 reviews) were included. Evidence was syn -
thesised narratively and not pooled, because methods and end -
points differed between studies. Its distribution by residue matrix
type and carrier family is shown in Figure 2 .
FIGURE 1
The recovery-to-application design chain.
Babatunde and de Souza 10.3389/fsufs.2026.2008486
Frontiers in Sustainable Food Systems 03 frontiersin.org
2 Residue streams and their bioactive
cargo
2.1 Scale of the resource and a matrix-based
classification
Fruit and vegetables are among the most loss-prone commodities
in the food system ( FAO, 2019 ; United Nations Environment
Programme, 2024). Industrial processing adds a concentrated and
predictable stream on top of these diffuse losses: citrus processing
waste represents roughly half of the fresh fruit mass (Suri et al., 2021),
grape pomace 20–30% of processed grapes ( Lopes et al., 2025 ),
mango processing 35–60% (w/w), or an estimated 14.7–25.2 million
tonnes per year (García-Mahecha et al., 2023), and tomato pomace
10–40% of processed tomatoes (Lu et al., 2022).
These streams are often organised by species, but matrix type is
more informative for encapsulation design because tissue structure
governs which compounds dominate, how they are bound and what
is co-extracted. Five recurring matrix types are peels, rinds and skins;
pomaces and marcs; seeds, kernels and stones; husks, hulls, shells and
pods; and leaves, stems and prunings (Table 1). This classification is
functionally useful rather than absolute: mixed streams such as
pomace can contain several tissues whose relative proportions
change with processing, so their extracts should not be assumed to
be compositionally equivalent.
2.2 Cargo classes
Polyphenols are the most widely studied bioactive compounds
recovered from fruit and vegetable processing residues, but neither
the term nor the extract describes a uniform cargo. In grape pomace,
skins carry hydroxycinnamic acids, flavonols and anthocyanins,
whereas seeds are rich in catechins and procyanidins and hold an
estimated 60–70% of the extractable phenolics ( Lopes et al., 2025).
A- and B-type proanthocyanidins are also distributed unevenly
among by-products. These structural differences affect polarity, pro-
tein binding and interfacial behaviour, and therefore influence
whether lipid-, protein- or polysaccharide-based carriers are suitable
(Lv et al., 2026). Carrier selection based only on total phenolic con -
tent is consequently unlikely to transfer reliably between residue
streams.
Pigments form a second important class, defined by colour per-
formance as well as chemistry. Anthocyanins, betalains, carotenoids
and chlorophylls can all be recovered from by-products, but they differ
in polarity and sensitivity to oxygen, light, pH and heat (Klavins et al.,
2026). The complexity of the starting material is illustrated by prickly
pear peel, from which 145 compounds, including 41 betalains and 16
carotenoids, were identified (Amaya-Cruz et al., 2019). Such mixtures
may offer complementary colour and antioxidant functions, but they
also make degradation pathways and marker selection more difficult
to interpret.
Essential oils and pectin are characteristic of citrus. Peels are the
largest citrus waste fraction and supply both essential oil, in which
d-limonene is the main monoterpene, and pectin at 20–30% of peel
(Suri et al., 2021; Koch et al., 2024), so the same peel can provide a
hydrophobic cargo and a polysaccharide wall material, the circularity
argument developed in Section 4.
Fibre-bound phenolics are less visible in conventional extract-
centred studies because ordinary solvent extraction largely misses
them. In lentil hulls, conjugated and bound phenolics occurred in
amounts similar to the free extractable fraction ( Sun et al., 2020 ),
while antioxidant activity in berry seed meals ranked insoluble-
bound > esterified > free ( Ayoub et al., 2016 ). These findings
agree that extractable phenolics alone understate the residue’s bio -
active potential. However, releasing bound fractions may require
chemical or enzymatic treatment that also changes molecular
form and co-extractives. The practical question is therefore not
simply how much phenolic material is present, but which fraction
can be recovered without compromising carrier compatibility,
safety or process feasibility.
FIGURE 2
Evidence map of the 298 included records by residue matrix type and carrier family. Each record was assigned to every matrix type and carrier family
named in its title or abstract, so a record can appear in more than one cell.
Babatunde and de Souza 10.3389/fsufs.2026.2008486
Frontiers in Sustainable Food Systems 04 frontiersin.org
2.3 Compositional variability as the defining
constraint
Variability, rather than scarcity, is the defining constraint for
residue-derived cargo. The literature consistently identifies culti -
var, tissue and maturity, processing stage, and stabilisation as
interacting sources of variation. Their effects are not interchange -
able: cultivar may dominate total composition, whereas drying or
storage can selectively alter individual compounds that determine
carrier interactions.
Cultivar. Across 17 Transylvanian grape pomaces, phenolic con-
centration and antioxidant activity differed markedly between culti-
vars (Chedea et al., 2025), and olive leaves from 32 cultivars ranged
from 4.92 to 18.29 mg/g dw in flavonoids and from 5.75 to 33.73 mg/g
dw in iridoids (Zhang C. et al., 2022).
Tissue and maturity. In pomegranate, the fruit fraction sepa -
rated flavonoid fingerprints more strongly than cultivar did ( Zhao
et al., 2022 ); gallotannins in mango seed kernel declined over
maturation in a cultivar-dependent way ( Alañón et al., 2021 ), and
green banana peel contained more phenolics than ripe peel
(Zhang J. et al., 2022 ).
Processing stage. By-products generated at different points of
the same process differ sharply. In olive oil production, leaves
were rich in flavonoids and terpenes but low in secoiridoids,
whereas alperujo carried phenolic alcohols and secoiridoids
(López-Salas et al., 2024 ).
Stabilisation. Drying and storage before extraction are not neutral
steps. Freeze-dried blueberry pomace retained the most antioxidant
components of three drying methods (Zhang Y . et al., 2025), and in
grape pomace, fraction-level phenolic composition was governed
mainly by cultivar while stabilisation produced strong compound-
level shifts, particularly among phenolic acids (Kiełtyka-Dadasiewicz
et al., 2026).
2.4 Implications for encapsulation design
Three consequences follow for encapsulation design. First, extract
composition should be treated as a variable input rather than a fixed
specification, because encapsulation efficiency, droplet size and stabil-
ity reported for one batch may not transfer to another cultivar, harvest
stage or drying route. Second, aggregate metrics such as total phenolic
content conceal compound-level differences that determine interfacial
behaviour and carrier compatibility ( Kiełtyka-Dadasiewicz et al.,
2026). Third, the same factors that change bioactive yield can change
contaminant and co-extractive profiles. Meaningful comparison
therefore requires a defined residue specification, compound-level
markers and contaminant screening before formulation begins.
3 Recovery and its downstream
consequences
3.1 Conventional versus green and
intensified extraction
Recovery has shifted from solvent maceration and Soxhlet extrac-
tion towards ultrasound-assisted extraction, microwave-assisted
extraction, supercritical CO₂ (SC-CO₂), pressurised liquid and sub-
critical water extraction, enzyme-assisted extraction, pulsed electric
fields and deep eutectic solvents, often in combination ( Floridas-
Galán and Paredes-Concepción, 2024; Jubril et al., 2025; Saini et al.,
2019). Intensified methods can shorten processing and reduce solvent
use: ultrasound with subcritical water produced 1.8-fold more poly-
phenols from Phyllanthus emblica pomace than solvent extraction,
manothermosonication reduced by 40-fold the time needed to obtain
TABLE 1 Residue matrix types, dominant bioactive cargo and features relevant to encapsulation.
Matrix type Typical streams Dominant cargo Encapsulation-relevant
features
References
Peels, rinds and skins Citrus, banana and plantain,
pomegranate, prickly pear,
tomato
Flavanone and flavonol
glycosides, anthocyanins,
betalains, carotenoids,
essential oils, pectin
Co-extracted pectin and essential oils;
pigment instability; strong effect of
maturity
Suri et al. (2021); Shehata et al.
(2021); Amaya-Cruz et al.
(2019); Zhang J. et al. (2022)
Pomaces and marcs Grape, olive (alperujo),
berry, cherry, tomato
Anthocyanins, procyanidins,
secoiridoids, lycopene, non-
extractable polyphenols
Mixed tissues (skin, seed, stem);
composition depends on cultivar and
stabilisation
Lopes et al. (2025); López-Salas
et al. (2024); Kiełtyka-
Dadasiewicz et al. (2026);
Domínguez-Rodríguez et al.
(2022)
Seeds, kernels and stones Mango kernel, grape seed,
berry seed meal
Gallotannins, mangiferin,
catechins, procyanidins,
bound phenolics
High phenolic density; cultivar and
maturation dependent
Alañón et al. (2021); García-
Mahecha et al. (2023); Ayoub et
al. (2016)
Husks, hulls, shells and
pods
Pulse hulls, coffee pulp and
peel
Conjugated and bound
phenolics, carotenoids,
tocopherols
Large bound fraction missed by
solvent extraction
Sun et al. (2020); Esquivel et al.
(2020)
Leaves, stems and
prunings
Olive leaves, cherry stalks Flavonoids, iridoids,
triterpenic acids
Wide inter-cultivar range Zhang C. et al. (2022);
Yüksekkaya et al. (2021); López-
Salas et al. (2024)
Babatunde and de Souza 10.3389/fsufs.2026.2008486
Frontiers in Sustainable Food Systems 05 frontiersin.org
equivalent antioxidant activity from pomegranate peel, and hybrid
ultrasound-microwave methods increased yields by up to 52% (Liang
et al., 2026; Güzel et al., 2020; Y ata-Franco et al., 2026). These gains
were obtained with different matrices, solvents and reference condi-
tions and are not directly comparable (Table 2).
A method described as “green” is not necessarily more selective,
scalable or compatible with the final carrier, nor necessarily lower in
impact: in a life-cycle assessment for apple pomace, supercritical CO₂
extraction emitted about 11 times more CO₂-eq per gram of bioactive
compounds than subcritical water extraction ( Bruna et al., 2025 ).
Recovery chemistry must match how the compound is held in the
matrix: pectinase liquefaction of carrot pomace raised β-carotene
extraction efficiency to about 90%, whereas bound phenolic acids in
cereal brans required alkaline hydrolysis rather than solvent extraction
alone (Jalali-Jivan et al., 2021; Guerrini et al., 2020 ). The apparent
advantage of a method may therefore reflect release of a particular
molecular fraction rather than a universally higher-quality extract.
Method selection should balance yield with molecular integrity, co-
extractive load, solvent residues and the requirements of the intended
food application.
3.2 Recovery outputs as encapsulation
inputs
An extract carries its recovery history into the carrier. Four attri-
butes are especially consequential: the solvent or emulsifier system
that remains, co-extracted matrix components, the molecular form of
the target compounds, and degradation caused by pretreatment. These
variables explain why two extracts described by the same total bioac-
tive content can form particles with different size, loading and stability.
Solvent. A pH-switchable hydrophobic natural deep eutectic sol-
vent (NADES) of decanoic and lauric acids recovered carotenoids
from orange peel and was used directly to prepare a 76.8 nm nano-
emulsion (Liu et al., 2026), and a supramolecular solvent of octanoic
acid extracted 96 to 100% of carrot peel carotenoids in 10 min at room
temperature before carrying them into nanostructured lipid carriers
(Muñiz-Bustamante et al., 2025).
Direct use of extraction media can simplify processing, but it also
shifts unresolved questions into the formulation. Many NADES con-
stituents, including organic acids, sugars, polyols and choline chloride,
are regarded as safe food components ( Zhou et al., 2023 ;
Cannavacciuolo et al., 2022). Nevertheless, component safety does not
by itself establish the safety or regulatory status of the final solvent
mixture at its delivered dose. High viscosity, variable composition,
limited techno-economic evidence and unclear clean-label status
remain important barriers (Gómez-Urios et al., 2026). Residual sol-
vent must therefore be treated as a formulation component that
requires specification, not as an invisible consequence of extraction.
When extraction and emulsification occur in one step, surfactant
choice sets both yield and particle properties: in a microemulsion-
subcritical water process for tomato peel, Tween 20 gave the highest
lycopene content whereas saponin gave the smallest particles
(14.6 nm) (Khanjani and Jafarizadeh-Malmiri, 2026).
Co-extractives can either support or disrupt carrier formation.
Proteins, polysaccharides and minerals in crude extracts are not auto-
matically impurities: an aqueous pomelo peel extract behaved as a
protein-pectin complex and assembled with quaternary ammonium
chitosan into a carrier for tea polyphenols (Ai et al., 2026). In contrast,
pomegranate peel ellagitannins bind food ingredients and salivary
glycoproteins, which can reduce functionality and cause astringency
(Andishmand et al., 2023). The direction of the effect depends on the
interaction: proteins and pectins can adsorb at interfaces or form elec-
trostatic complexes, whereas tannins cross-link proteins through
hydrogen bonding and hydrophobic interactions. The strong depen-
dence of poly(lactide-co-glycolide) nanoparticle properties on the
tropical fruit extract rather than polymer grade further shows that the
crude matrix can control formulation behaviour (Silva et al., 2014).
Purifying every extract is not always necessary, but its major co-
extractives must be characterised and tested as functional variables.
Molecular form and pre-treatment. Highly polymerised grape
seed tannins show low oral bioavailability; microwave-assisted depo-
lymerisation with gallic acid degraded 99.8% of the polymer to mainly
epicatechin monomers, which were then encapsulated in phosphati-
dylcholine liposomes with 83.1% efficiency (Morales et al., 2025).
3.3 Purification level as a design decision
Purification can improve compositional control but does not pro-
vide a universal optimum. Resin enrichment increased flavonoid con-
centration in acerola pomace extract up to 20-fold, after which
nanocapsules reached 81.4% encapsulation efficiency (Freitas et al.,
2025). More defined systems were obtained with purified trans-resve-
ratrol in lecithin nanoparticles below 160 nm and purified pomegran-
ate peel ellagitannins in zein nanoconjugates with 83.2% binding
efficiency (Karagül and Uğraş, 2025; Foujdar et al., 2023). These results
support purification when a defined marker and dose are required, but
they do not show that purified fractions outperform crude extracts in
food function, where co-extracted compounds may contribute syner-
gistic activity or physical stability.
The appropriate purification level is therefore a design decision
rather than a fixed hierarchy. Greater purification improves reproduc-
ibility and simplifies carrier optimisation, but adds cost, solvent use
and yield loss and may remove useful co-extractives. Conversely,
crude extracts preserve chemical breadth but increase batch variability
and the risk of carrier interference. Even clarification can determine
whether a nanoscale system forms at all (Luca et al., 2012; Silva et al.,
2014). Purification should consequently be justified against a target
specification for composition, payload, food performance and safety.
3.4 Integrated extraction-encapsulation and
cascade processing
Two integration strategies are emerging. One couples extraction
and encapsulation within the same process line, as in liposome-inte-
grated ultrasound and subcritical water extraction, SC-CO₂ recovery
followed by nanoliposome formation, and supercritical assisted atomi-
sation of olive pomace extract with maltodextrin (Liang et al., 2026;
Sheikhzadeh et al., 2026; Aliakbarian et al., 2017). The second uses
cascade processing to recover several fractions sequentially. Both
approaches may reduce handling and solvent-removal steps, but pro-
cess integration also couples failures: a solvent or surfactant chosen to
maximise extraction can restrict carrier selection, food compatibility
or regulatory acceptance. Table 2 therefore compares recovery routes
on consistent criteria (properties transferred downstream, solvent and
energy demand, scale readiness), not by yield alone.
Babatunde and de Souza 10.3389/fsufs.2026.2008486
Frontiers in Sustainable Food Systems 06 frontiersin.org
TABLE 2 Recovery routes, extract attributes and implications for carrier choice.
Recovery route Residue
examples
Extract attributes
relevant to
encapsulation
Solvent and
energy
demand; scale
readiness
Implication for
carrier
References
Ultrasound-assisted, alone or
coupled (subcritical water,
manothermosonication, cold
plasma)
Phyllanthus emblica
pomace,
pomegranate peel,
mango peel, olive
leaf, orange peel
Hydroalcoholic or aqueous
crude extract; short time
and mild temperature
preserve thermolabile
compounds
Low solvent and
energy; pilot and
industrial units exist
Compatible with most
carriers after solvent
removal; high co-
extractive load favours
carriers judged on total
phenolics
Liang et al. (2026);
Güzel et al. (2020);
Khalili et al. (2025);
Saborirad et al. (2024);
Jubril et al. (2026b);
Prevete et al. (2024)
Supercritical CO₂ and
supercritical assisted
atomisation
Pumpkin peel,
pomegranate peel,
grape pomace, olive
pomace
Solvent-free, lipophilic-rich
extract; may yield less total
polyphenol than
hydroalcoholic extraction
No organic solvent;
high pressure and
capital cost;
industrial for oils
Suits lipid carriers
(nanoliposomes) or
direct atomisation with
maltodextrin
Sheikhzadeh et al.
(2026); Zahed and
Kenari (2024); Hałdys
et al. (2026);
Aliakbarian et al. (2017)
Hydrophobic NADES and
supramolecular solvents
Orange peel, carrot
peel, taperebá peel
Solvent retained with
extract; high carotenoid
recovery at low temperature
Solvent stays in
product; low energy;
laboratory
Solvent becomes oil
phase or lipid carrier
component; avoids
solvent removal step
Liu et al. (2026);
Muñiz-Bustamante et
al. (2025); Mello et al.
(2025); Della Posta et al.
(2022); Gómez-Urios et
al. (2026)
Hydrophilic NADES Olive pomace High phenolic yield;
residual NADES in extract
Viscous solvent to
remove or tolerate;
laboratory
Carrier must tolerate or
remove NADES;
nanoemulsion
incorporation reported
Chanioti et al. (2021)
Enzyme-assisted (pectinase,
cellulase)
Carrot pomace,
defatted grape seed
Releases matrix-bound
compounds; hydrolysed
polysaccharides co-
extracted
Aqueous; enzyme
cost; industrial in
juice processing
Extraction into
nanoemulsion micelles
possible; co-extracted
sugars may affect drying
Jalali-Jivan et al. (2021);
Sun et al. (2024)
Extraction in carrier-forming
solutions
Acerola by-products
(chitosan), grape
pomace (gliadin-rich
solvent)
Extract and carrier material
combined from the start
Low solvent;
laboratory
One-step route to
nanoparticles; reduced
solvent use
Silva et al. (2021);
Carpentieri et al. (2023)
Microemulsion or
simultaneous extraction-
nanoemulsification
Tomato peel and
processing waste
Surfactant choice sets both
yield and particle size
Surfactant-intensive;
laboratory
Emulsifier should be
selected for the target
food and label, not only
for yield
Khanjani and
Jafarizadeh-Malmiri
(2026); Makou et al.
(2024)
Adsorbent resin enrichment or
purification
Acerola pomace,
grape seed, grape
pomace,
pomegranate peel
Enriched fraction or isolated
compound with defined
composition
Extra solvent for
elution; industrial for
some extracts
Enables carriers
optimised for one
molecule; higher cost
and yield loss
Freitas et al. (2025);
Fathi et al. (2023);
Karagül and Uğraş
(2025); Foujdar et al.
(2023)
Chemical modification before
encapsulation
(depolymerisation)
Grape seed
condensed tannins
Lower molecular weight,
reduced protein binding
Added reagent and
energy; laboratory
Improves liposome
loading and
bioaccessibility
Morales et al. (2025)
Clarification (removal of
insoluble fines)
Sour cherry pomace Lower particle load; smaller
dispersed particles
Low demand; readily
scalable
Nano-suspension
formed only after
clarification; EE rose to
about 98%
Luca et al. (2012)
Ratings are qualitative, from the cited studies; yields obtained under different conditions are not directly comparable. EE, encapsulation efficiency; NADES, natural deep eutectic solvent.
Babatunde and de Souza 10.3389/fsufs.2026.2008486
Frontiers in Sustainable Food Systems 07 frontiersin.org
4 Carrier systems and the selection
logic
Carrier choice is often treated as a formulation exercise that
begins after extraction, yet the recovered material already constrains
what can be built. Food-grade selection strategies commonly classify
carriers by cargo hydrophilicity and loading mechanism (Huang et al.,
2024), reflecting the broader principle of chemical compatibility
between the bioactive and the carrier (Pateiro et al., 2021; Assadpour,
2018). For residue extracts, compatibility must also include co-extrac-
tives, residual solvent, target-food conditions and the required release
profile. No carrier family is inherently superior across these dimen-
sions; Table 3 compares them by cargo fit, tolerance of crude extracts,
food-matrix compatibility, regulatory constraints and cost.
4.1 Carrier families
Protein-based carriers can combine food-grade status with strong
interfacial activity, but their performance depends on both protein
source and extract composition. Zein nanoparticles formed by anti-
solvent precipitation loaded grape pomace extract with 89–97%
encapsulation efficiency, whereas whey protein concentrate nanopar-
ticles retained only 14.4–27.9% of mandarin peel extracts ( Luque-
Alcaraz et al., 2025 ; Hu et al., 2019 ). This contrast may reflect
differences in cargo polarity, marker definition and formation route
rather than a simple zein advantage. Animal proteins are often better
matched to lipophilic compounds than plant proteins, but allergenic-
ity, solubility, pH sensitivity and labelling can outweigh loading per-
formance in a specific food (Oliveira et al., 2025a; Y an et al., 2022).
Polysaccharide-based carriers are most persuasive when con -
trolled release or pH-responsive protection is required. Pectin
nanoparticles reduced release of mandarin peel flavonoids in simu-
lated gastric fluid from 73% for the free extract to 28.8%, while chito-
san-alginate nanoparticles retained 72–78% of a grape-seed
procyanidin under gastrointestinal conditions (Hu et al., 2017; Wu et
al., 2023). Seed gums broaden the range of residue-compatible wall
materials (Akcicek et al., 2021; Khalili et al., 2025). However, delayed
release is beneficial only if the compound is subsequently liberated
and solubilised in the intestine; gastric retention alone cannot estab-
lish improved bioavailability.
Lipid-based carriers are well suited to amphiphilic compounds,
but improvements in one metric may compromise another. Grape-
seed proanthocyanidin liposomes exceeded 90% encapsulation effi-
ciency and pomegranate peel nanophytosomes reached 95.0% at
144–167 nm (López-Polo et al., 2026; Neslihan Dundar et al., 2023).
Layer-by-layer coating of apple peel polyphenol liposomes increased
encapsulation from 86.5 to 92.0%, but also increased particle size from
267 to 422 nm (Hu et al., 2025). The additional layer may improve
protection, yet larger size, processing complexity and altered digestion
must be evaluated before the higher encapsulation value is treated as
better overall performance.
Particle-stabilised emulsions require careful terminology and per-
formance criteria. Tea seed cake protein particles reduced droplet size
from 337 to 89 μm as particle concentration increased, so the
nanoscale element was the interfacial particle rather than the emul-
sion droplet (Liang et al., 2022). Genuine Pickering stabilisation, in
which particles adsorb at the oil–water interface, should also be dis-
tinguished from systems in which plant particles mainly thicken or
structure the continuous phase (Sarkar et al., 2020b ). By-product-
derived particles may provide robust interfaces and support circular-
ity, but opacity, sedimentation, mouthfeel and digestibility can limit
their use in foods even when the emulsion is physically stable (Muiz
et al., 2023).
Electrospun and electrosprayed structures. Electrospinning
encapsulates extracts in fibres without heat: gelatin nanofibres pro-
tected pomegranate peel extract during simulated digestion (Saadat et
al., 2021), and citric acid cross-linked gelatin nanofibres reached
83.7% loading efficiency for tomato pomace phenolics ( Ince et
al., 2026).
4.2 Formation routes and tolerance for
crude extracts
Carriers do not simply protect a crude extract; they can fraction-
ate it and change the delivered composition. When olive leaf extract
was loaded into three phospholipid matrices, one retained non-polar
triterpenes preferentially while the others favoured polar phenolics
(Baljak et al., 2025 ). The encapsulated fraction therefore differed
chemically from the starting extract. A second complication arises
when the carrier is itself bioactive: chitosan coatings enriched with
pineapple by-product phenolics reduced liposome encapsulation effi-
ciency from 66% to 36–53% while tripling total phenolics (Bassan et
al., 2025). These findings show why marker-specific encapsulation
efficiency must be interpreted together with payload and post-encap-
sulation composition.
Formation route adds a second layer of constraints. Antisolvent
precipitation of zein or gliadin requires an ethanol-compatible extract,
ionic gelation generally favours polar cargo, thin-film hydration uses
organic solvent, and high-pressure or ultrasonic emulsification intro-
duces energy and heat. Drying can then change the system that was
initially characterised; freeze-drying increased liposome size from
76–81 nm to 199–283 nm ( Marín et al., 2018 ). Carrier selection
should therefore include the complete manufacturing sequence,
because performance measured before drying may not represent the
ingredient added to food.
4.3 A decision framework for carrier
selection
Figure 3 organises carrier selection as a sequence of five questions,
each answered from information that should be available before for-
mulation begins.
(1) What is the polarity and molecular form of the cargo?
Hydrophobic carotenoids and essential oils point to nano -
emulsions, nanostructured lipid carriers, Pickering emulsions
or zein (Liu et al., 2026; Gao et al., 2023); polar phenolics and
anthocyanins point to ionic gelation, protein-polyphenol com-
plexes, liposomes or phytosomes (Han et al., 2015; Fang et al.,
2024); and polymeric tannins may need depolymerisation first
(Morales et al., 2025).
(2) How pure is the extract? Crude extracts favour carriers that
tolerate mixtures and are judged on total load rather than one
marker; enriched fractions justify carriers optimised for a
defined molecule ( Freitas et al., 2025 ; Karagül and
Uğraş, 2025).
Babatunde and de Souza 10.3389/fsufs.2026.2008486
Frontiers in Sustainable Food Systems 08 frontiersin.org
TABLE 3 Critical comparison of carrier families for residue-derived bioactives.
Carrier family Common
formation
routes
Cargo fit Performance
with crude
extracts and
limitations
Food fit;
regulatory
and cost
Residue-
derived
versions
Key sources
Protein nanoparticles
and complexes (zein,
gliadin, whey, pea,
seed proteins)
Antisolvent
precipitation, pH-
driven assembly,
complexation, spray
drying
Hydrophobic and
amphiphilic
phenolics,
carotenoids
High EE with grape
pomace extract (89 to
97%) but low EE with
some citrus peel extracts
(14 to 28%)
Wide; allergen
labelling (milk, soy,
gluten); low to
moderate cost
Apple seed,
camellia seed cake,
tea residue,
rapeseed meal,
pumpkin leaf
proteins; oilseed
cake nanofibrils;
pea protein-potato
peel starch
coacervates
Luque-Alcaraz et
al. (2025); Hu et al.
(2019); Gani et al.
(2022); Yu et al.
(2022); Zeng et al.
(2026); Lan et al.
(2021); Salević-
Jelić et al. (2025);
Saricaoglu et al.
(2026); Akram et
al. (2025)
Polysaccharide
nanoparticles
(chitosan, alginate,
pectin, gums)
Ionic gelation,
polyelectrolyte
complexation, coating
Polar phenolics,
anthocyanins
Widely used with crude
hydroalcoholic extracts;
controlled gastric release
Acidic beverages,
coatings; chitosan
status varies by
jurisdiction; low cost
Peel pectins
(passion fruit,
Akebia, banana,
hawthorn pomace)
Han et al. (2015);
Hu et al. (2017);
Wu et al. (2023); Li
et al. (2024); Cai et
al. (2020); Méndez
and López (2020);
Jiang et al. (2020)
Liposomes,
phytosomes,
niosomes and
proniosomes
Thin-film hydration,
reverse evaporation,
slurry method
Amphiphilic and
polar phenolics
Carriers select within
extracts; coatings change
EE and total load
Beverages, dairy;
lecithin source
labelling; moderate
to high cost, solvent-
based preparation
Pomelo peel
protein-pectin
coatings
López-Polo et al.
(2026); Neslihan
Dundar et al.
(2023); Hu et al.
(2025); Baljak et al.
(2025); Bassan et
al. (2025); Ilgaz et
al. (2025)
Nanoemulsions, NLC
and self-emulsifying
systems
High-energy
(ultrasound,
homogenisation) or
low-energy
emulsification
Carotenoids,
essential oils,
lipophilic extracts
Solvent and emulsifier
choice set yield and size;
oil type affects
performance
Beverages, dairy;
synthetic surfactants
conflict with clean
label; low to
moderate cost
Lignin and bael
shell emulsifiers
Liu et al. (2026);
Muñiz-Bustamante
et al. (2025);
Chiang and
Chiang (2026);
Pinna et al. (2025);
Mukhtar et al.
(2026); Sharma et
al. (2025); Katsouli
et al. (2024)
Pickering emulsions Particle adsorption at
oil–water interface
Hydrophobic
cargo
Residual lignin and
hemicellulose can
improve performance
Opaque foods,
spreads; residue
particles may need
safety data; low cost
Bran and okara
CNF , pomelo peel
nanocellulose,
apple pomace
nanofibrils,
coconut pomace
cellulose; litchi peel
CNC, maple leaf
LCNC, waste CNF
Chen et al. (2025);
Gao et al. (2023);
Zhang C. et al.
(2024); Wang et al.
(2025); Liang et al.
(2022); Ma et al.
(2023); Ji and
Wang (2024);
Kanai and
Kawamura (2025)
(Continued)
Babatunde and de Souza 10.3389/fsufs.2026.2008486
Frontiers in Sustainable Food Systems 09 frontiersin.org
(3) What is the target matrix? Clear acidic beverages, emulsified
dairy products, heated bakery products and polymer films
impose different requirements on particle charge, size, heat
stability and compatibility (Section 6).
(4) How severe is processing and storage? Thermal treatment, low
pH, ionic strength and light should be matched to carriers
shown to survive them, for example quaternary ammonium
chitosan-pomelo peel carriers that retained 58.9% of tea poly-
phenols after 12 h of UV exposure against 43.0% for the free
compounds (Ai et al., 2026).
(5) What label and regulatory constraints apply? Synthetic surfac-
tants such as polysorbates can be replaced by saponin
(Khanjani and Jafarizadeh-Malmiri, 2026), residue-derived
lignin (Mukhtar et al., 2026) or pectin-containing lignocellu-
losic nanofibrils from apple pomace, proposed as a clean-label
emulsifier (Zhang C. et al., 2024).
This framework is necessarily qualitative because direct carrier
comparisons using the same extract, dose and stress conditions
remain uncommon. Where such comparisons are available, the
ranking changes with the endpoint: a system that gives smaller par -
ticles may not provide the highest loading, retention or bioacces -
sibility ( Freitas et al., 2025 ; Baljak et al., 2025 ; Lu et al., 2020 ).
Figure 3 therefore places the target food and the required
TABLE 3 (Continued)
Carrier family Common
formation
routes
Cargo fit Performance
with crude
extracts and
limitations
Food fit;
regulatory
and cost
Residue-
derived
versions
Key sources
Electrospun and
electrosprayed fibres
and particles
Electrospinning,
solution blow
spinning
Heat-sensitive
phenolics,
carotenoids
Loading efficiency up to
84% for tomato pomace
phenolics
Coatings, packaging;
low throughput
Cellulose
nanowhisker
reinforcement
Saadat et al.
(2021);
Paczkowska-
Walendowska et al.
(2023); İnanç
Horuz and
Belibağlı (2019b);
Ince et al. (2026)
Whole residue
powders and flours as
wall materials
Homogenisation,
freeze-drying
Oils and
carotenoids
High EE (93 to 96%)
and storage stability
Opaque or dry
foods; contaminant
screening needed;
very low cost
Guava peel and
seed, pitaya peel,
jackfruit seed flour,
peanut oil-cake;
orange juice by-
product
Schappo et al.
(2026); Andrade et
al. (2026); Oliveira
et al. (2025b);
Rebeca et al.
(2026); Kaderides
et al. (2019)
CNC, cellulose nanocrystals; CNF , cellulose nanofibrils; EE, encapsulation efficiency; LCNC, lignin-containing CNC; NLC, nanostructured lipid carriers.
FIGURE 3
Carrier-selection decision map structured as a sequence of five questions.
Babatunde and de Souza 10.3389/fsufs.2026.2008486
Frontiers in Sustainable Food Systems 10 frontiersin.org
performance metric before final carrier selection. The available
findings support compatibility-based screening, but not a universal
ranking of carrier families.
4.4 Circularity: residue-derived wall
materials
Studies show that residues can supply both the bioactive cargo
and part or all of the carrier material. Four main classes recur:
whole residue powders and flours, proteins and peptides, pectins,
and cellulose- or lignin-based particles. This strategy strengthens
material circularity, but it also transfers natural variability, pos -
sible allergens and contaminants from a second residue-derived
input into the formulation. Compared to the same criteria, the
classes differ. Whole powders need the least processing and reach
EE above 93% ( Andrade et al., 2026 ; Oliveira et al., 2025b ), but
their active components are undefine and they carry the residue’s
full contaminant load. Proteins add interfacial and film-forming
function ( Gani et al., 2022 ; Lan et al., 2021 ) but bring allergen,
off-flavour and solubility risks. Pectins are established food ingre -
dients, but their carrier function depends on degree of esterifica -
tion ( Méndez and López, 2020 ). Cellulose and lignin particles
form robust Pickering interfaces ( Kanai and Kawamura, 2025 ; Ma
et al., 2023 ), yet can increase opacity and alter texture and digest -
ibility ( Muiz et al., 2023 ), and lignin lacks an established food-use
statu. Whole residue powders and flours minimise fractionation
and can provide several stabilising components at once. Freeze-
dried red pitaya peel encapsulated hybrid crude palm oil with
95.6% efficiency and 60-day stability at 4 °C, while jackfruit seed
flour carried the same oil at 93.2% efficiency ( Andrade et al.,
2026; Oliveira et al., 2025b ). Orange-juice processing residue also
improved retention of pomegranate peel phenolics relative to the
free extract, although substantial degradation still occurred
during baking ( Kaderides et al., 2019 ). These systems demonstrate
feasibility, but high encapsulation values do not reveal which resi -
due components provide the barrier or how batch variability will
affect performance.
Proteins and peptides recovered from seeds, cakes and leaves offer
interfacial and film-forming functions alongside circularity. Apple
seed protein, rapeseed meal peptides and pumpkin leaf protein have
been assembled into carriers, while protein nanofibrils from oilseed
cakes are emerging as a further class (Gani et al., 2022; Lan et al., 2021;
Salević-Jelić et al., 2025; Saricaoglu et al., 2026). Their practical advan-
tage depends on achieving reproducible solubility and functionality
after recovery. Allergenicity, off-flavour and source-dependent com-
position remain more important constraints than laboratory particle
formation alone.
Residue-derived pectins illustrate how recovery and carrier func-
tion are inseparable. Passion fruit peel pectin formed zein-caseinate-
pectin nanoparticles for curcumin, and Akebia trifoliata peel pectin
coated curcumin-loaded zein nanoparticles (Li et al., 2024; Cai et al.,
2020). Banana peel pectin formed polyelectrolyte nanoparticles with
amphiphilic chitosan only when its degree of esterification was low
(24.3%) (Méndez and López, 2020). The decisive variable was there-
fore not simply that the material was pectin, but the molecular proper-
ties created during extraction. Residue-derived pectins require
functional specifications comparable to those used for conventional
commercial grades.
Cellulose- and lignin-based particles can create strong inter -
facial barriers, but their behaviour depends on particle dimen -
sions, surface chemistry and residual non-cellulosic components.
Waste-derived cellulose nanofibres have stabilised Pickering
emulsions comparably to wood-derived fibres, litchi-peel cellulose
nanocrystals produced stable emulsions above 0.7 wt% across
changes in pH, ionic strength and temperature, and lignin-con -
taining cellulose nanocrystals from maple leaves acted as natural
stabilisers ( Kanai and Kawamura, 2025 ; Ma et al., 2023 ; Ji and
Wang, 2024). Pomelo-peel nanocellulose and sugarcane-bagasse
lignin nanoparticles extend this approach ( Gao et al., 2023 ;
Mukhtar et al., 2026 ). Y et stability in simplified emulsions does
not establish acceptable texture, colour, digestibility or regulatory
status in foods.
Taken together, residue-derived carriers make full-loop for -
mulations scientifically plausible, including systems in which
guava residues supply several formulation components and cas -
cade biorefineries recover both bioactives and structural polymers
from pitahaya, orange and potato peels ( Schappo et al., 2026 ;
Arroyo-Esquivel et al., 2025 ; Espinosa et al., 2022 ; Li et al., 2026 ).
The evidence nevertheless remains mainly proof of concept.
Residue use or high EE alone does not demonstrate circularity or
environmental benefit, because intensive extraction, hydrolysis
and drying can offset it: in a comparative life-cycle assessment of
phenolic extracts, electricity for freeze-drying and chemical treat -
ment dominated impacts, and a cascade route had the highest
footprint ( Carpentieri et al., 2025 ). Circularity claims therefore
require mass balance, consistent carrier specifications, contami -
nant control, and life-cycle assessment (LCA) and techno-eco -
nomic analysis (TEA) against conventional wall materials.
5 Characterisation and the
performance-metric problem
5.1 Standard descriptors
Particle or droplet size, polydispersity index, zeta potential, elec-
tron microscopy and infrared spectroscopy are standard descriptors
used to infer dispersion quality and cargo-carrier interactions (Luque-
Alcaraz et al., 2025; Foujdar et al., 2023). They are necessary, but they
are not direct measures of protection, delivery or food performance
(Jubril et al., 2026a). Each has limits: dynamic light scattering gives an
intensity-weighted diameter biased towards large particles; the poly-
dispersity index does not resolve multimodal populations; zeta poten-
tial depends on the pH and ionic strength of the medium; microscopy
samples few particles, often after drying; and infrared spectroscopy
indicates interactions qualitatively but does not quantify encapsula-
tion. Size measured in water immediately after preparation may differ
substantially after drying, storage or addition to food; freeze-drying,
for example, roughly tripled liposome diameter (Marín et al., 2018).
Characterisation should therefore be repeated after processing, drying,
storage, food incorporation and simulated digestion.
Babatunde and de Souza 10.3389/fsufs.2026.2008486
Frontiers in Sustainable Food Systems 11 frontiersin.org
5.2 Encapsulation efficiency, payload and
retention
EE, the fraction of added cargo associated with the carrier, varies
widely across residue-derived systems, from 14.4% for whey protein
concentrate carrying mandarin peel extract to above 95% for pitaya-
peel material carrying palm oil (Hu et al., 2019; Andrade et al., 2026).
Reviews often report values above 90% for protein and polysaccharide
matrices (Y ata-Franco et al., 2026). This range reflects differences in
cargo, carrier, marker, separation method and calculation as much as
formulation quality. It should be distinguished from loading capacity
(payload; cargo per mass of carrier), retention (cargo remaining after
a defined stress), release (cargo liberated over time into a defined
medium), bioaccessibility (cargo released and solubilised after diges-
tion) and bioavailability (cargo absorbed and reaching circulation;
Table 4). Values obtained by different methods should not be used to
rank formulations. Three recurring limitations prevent EE from serv-
ing as a stand-alone performance measure.
(1) EE is measured against different markers, some studies using
total phenolic content and others a single compound such as
epicatechin (Carpentieri et al., 2023) or oleuropein (Baljak et
al., 2025 ). Because carriers select within crude extracts
(Section 4.2), EE for one marker does not describe the whole
extract, and total phenolic assays can hide compound-level
shifts (Kiełtyka-Dadasiewicz et al., 2026).
(2) EE says nothing about payload. Camellia seed cake albumin
nanoparticles had an EE of 43.8% but an effective loading of
only 5.9% ( Yu et al., 2022 ), and apple seed protein
nanocapsules raised EE from 69 to 80% ( Gani et al., 2022 )
without indicating how much carrier is needed per dose. A
high EE at low loading may require impractical amounts of
carrier in a food.
(3) EE can move in the opposite direction to delivered amount: in
chitosan-coated liposomes, EE fell while total phenolic con-
tent tripled because the coating carried additional phenolics
(Bassan et al., 2025). Size and EE can also trade off (Freitas et
al., 2025; Khanjani and Jafarizadeh-Malmiri, 2026).
Retention under relevant stress provides a more useful comple-
ment to EE. β-Carotene retention was about 25% under UV light and
32% at pH 4 in nanoemulsions, compared with complete loss from the
ethanolic extract (Jalali-Jivan et al., 2021). Even these values must be
interpreted against the target food and expected exposure time. Some
application studies omit size, zeta potential and EE altogether
(Rasheed, and Jum’a, 2026), so their outcomes cannot be attributed to
nanoscale delivery.
5.3 Food-relevant stability, digestion and
bioaccessibility
In vitro digestion is the main bridge between carrier characteri -
sation and a plausible nutritional benefit. Nanostructured lipid car -
riers raised the bioaccessibility of free xanthophylls and β-carotene
from pumpkin peel to about 65% and that of esterified carotenoids
from 3 to 20%, while pectin nanoparticles delayed flavonoid release
in simulated gastric fluid ( Pinna et al., 2025 ; Hu et al., 2017 ).
Emerging work also examines intestinal permeability and gut
TABLE 4 Performance metrics, interpretive limitations and minimum reporting requirements.
Metric What it shows Common interpretive limitation Minimum reporting
Particle or droplet size
and PDI
Physical scale and uniformity “Nano” applied to systems of several hundred nm
or to micrometre droplets; size changes after
drying not reported
Method, dispersion medium, number and
intensity distributions; size before and after
drying and in the food
Zeta potential Electrostatic stability Measured only in water at one pH pH and ionic strength of measurement; values at
food pH
Microscopy and FTIR Morphology; cargo-carrier
interactions
Few particles imaged, often dried; FTIR is
qualitative
Sample preparation; pair with quantitative size
and EE
Encapsulation
efficiency
Encapsulated cargo as % of cargo
added
Different markers (total phenolics or single
compound); ignores carrier-borne actives
Marker compound and method; EE for total and
key compounds
Loading capacity
(payload)
Encapsulated cargo as % of carrier
(particle) mass
Rarely reported; high EE may coincide with low
loading
Loading capacity and carrier dose needed per
serving
Retention under stress Cargo remaining after a defined
stress, % of initial
Stress conditions unrelated to the target food Retention under heat, pH, ions, light and storage
matching the food
In vitro release Cargo liberated over time into a
defined medium
Media and sink conditions unrelated to food or
gut
Medium, temperature, sampling method and
kinetic model
Bioaccessibility Fraction released and solubilised
(e.g., micellar phase) after in vitro
digestion
Results differ between static and dynamic models
and with oil, emulsifier and calcium levels
Named model and protocol (for example
INFOGEST); oil, emulsifier, calcium and bile
levels; controls with free extract
Bioavailability Fraction absorbed and reaching
circulation
Inferred from in vitro bioaccessibility Cell, animal or human data; state the model
Antioxidant or
antimicrobial activity
Functional potential Chemical assays may favour free extract; poor
proxy for food function
Activity in the food matrix with free-extract and
synthetic controls
EE, encapsulation efficiency; FTIR, Fourier-transform infrared spectroscopy; PDI, polydispersity index.
Babatunde and de Souza 10.3389/fsufs.2026.2008486
Frontiers in Sustainable Food Systems 12 frontiersin.org
microbial responses ( Miyagi et al., 2026 ; Fierri et al., 2026 ). These
findings support altered gastrointestinal behaviour, but delayed gas -
tric release, micellarisation, epithelial uptake and systemic bioavail -
ability are distinct endpoints and should not be treated as
interchangeable evidence of nutritional benefit. In vitro digestion
measures bioaccessibility only; absorption requires cell, animal or
human models, and human data for residue-derived nanocarriers
remain rare.
The digestion model can change both the magnitude and inter -
pretation of an apparent benefit. Apple pomace showed polyphenol
bioaccessibility similar to whole apple in a semi-dynamic model,
but variability was much higher (coefficient of variation 69.4%
versus 19.5%) ( Lopez-Rodulfo et al., 2025 ). For aged citrus-peel
polymethoxyflavones, nanoemulsions and Pickering emulsions
produced about 14-fold higher bioaccessibility than bulk oil in a
pH-stat model but only two- to fourfold increases in the dynamic
TIM-1 system ( Lu et al., 2020 ). Evidence is therefore consistent
that formulation changes digestion, but divided on the size of the
benefit. Every claim should specify the model, matrix and
endpoint.
Formulation variables that are not reported consistently can also
reverse performance. In standardised digestion of β-carotene emul-
sions, bioaccessibility peaked at 93.2% at 10% oil and declined at 20%
oil, fell from 65.5 to 23.7% as calcium increased from 0.525 to 10 mM,
and varied with emulsifier because caseinate-stabilised droplets floc-
culated in the gastric phase (Tan et al., 2020b; Tan et al., 2019; Tan et
al., 2020a). High surfactant concentrations also inhibited lipid hydro-
lysis (Salvia-Trujillo et al., 2019). Smaller droplets can increase inter-
facial area, but they do not automatically improve bioaccessibility
when digestion is constrained by mineral bridging, flocculation or
inhibited lipolysis.
Residue studies confirm the influence of wall and food. Inulin in
the wall of microparticles carrying mango by-product carotenoids
enhanced β-carotene bioaccessibility almost 2-fold (Cabezas-Terán et
al., 2025), and co-digestion with milk gave bioaccessibility of 8.8 to
75.5%, highest with whole milk (Cabezas-Terán et al., 2025). Lycopene
from tomato by-products degraded during spray-drying microencap-
sulation (Szabo et al., 2021), whereas high-pressure homogenisation
improved the bioaccessibility of papaya by-product carotenoids (Lara-
Abia et al., 2023).
5.4 Reporting inconsistencies and a
minimum reporting set
Variation in markers, media and digestion protocols prevents reli-
able ranking of carriers across publications and makes formal meta-
analysis premature. Table 4 proposes a minimum reporting set that
links each descriptor to the conditions needed for interpretation.
Digestion studies should use the harmonised INFOGEST static pro-
tocol, or its semi-dynamic version when gastric emptying matters, so
that data are comparable between laboratories, and should state any
deviation, oil and emulsifier content, calcium and bile concentrations,
and include a free-extract control ( Brodkorb et al., 2019 ; Mulet-
Cabero et al., 2020; Tan et al., 2019). Harmonised reporting will not
remove biological variability, but it will separate genuine carrier effects
from methodological differences.
6 Food applications
6.1 Application classes
Food applications span beverages, dairy and bakery products,
meat and fish, edible oils, fresh-produce coatings, and active or intel-
ligent packaging (Table 5). The evidence is most convincing when
formulation performance is measured after incorporation into the
food and over realistic storage. Many reports instead infer applicabil-
ity from particle characterisation or short model-system tests. The
sections below therefore distinguish proof of formulation from evi-
dence that the carrier improves a real product. All studies in Table 5
were laboratory scale; none reported pilot or industrial validation.
Beverages expose carriers to low pH, dissolved minerals, dilution
and optical constraints. Nanoemulsions containing olive-pomace phe-
nolics and tomato-pomace lycopene showed that carrier oil influenced
bioaccessibility, while a β-carotene nanoemulsion from Citrus reticu-
lata peel intensified juice colour and increased in vitro bioaccessibility
(Katsouli et al., 2024; Barman et al., 2020). These results support bev-
erage use for lipophilic pigments, but colour enhancement and diges-
tive release do not by themselves establish long-term colloidal stability
or consumer acceptance.
Dairy matrices can protect bioactives through fat and protein
interactions, but those same interactions can obscure the contribution
of the carrier. Pomegranate- and pomelo-peel coatings extended the
shelf life of Domiati cheese and paneer, while microencapsulated
grape-pomace extract remained stable over 90 days and increased anti-
oxidant activity and intestinal bioaccessibility in Greek-style yogurt
(El-Ashaal et al., 2025; Joshi et al., 2025; López-Astorga et al., 2025).
The findings support dairy as a promising matrix, but direct compari-
sons across fat levels, protein structures and unencapsulated controls
are needed to separate carrier effects from matrix-assisted protection.
Bakery products provide a stringent test because heat, low water
activity and dough interactions can degrade cargo or immobilise it.
Orange-peel extract encapsulated in maltodextrin and gum arabic pro-
duced cakes without significant sensory differences, and chitosan-tri-
polyphosphate particles carrying grape, jaboticaba and hibiscus
by-product extracts retained more than 89% of phenolics in tapioca
heated to 200 °C ( Mahmoud et al., 2016 ; Oliveira et al., 2026 ).
Encapsulated citrus flavonoids in chocolate cookies were also not dis-
tinguished from controls, although the chocolate matrix may have
masked sensory changes (Dellapina et al., 2025; Indiarto et al., 2022).
Thermal retention is encouraging, but retained analytical content must
be linked to release during digestion before nutritional claims are made.
Meat and fish applications provide stronger functional endpoints
because oxidation and microbial growth can be measured during stor-
age. Pomegranate-peel nanoparticles reduced peroxide values, thio-
barbituric acid reactive substances and total volatile basic nitrogen in
meatballs over 15 days at 4 °C while maintaining acceptable colour
and odour; liposomes carrying residue extracts also maintained squid-
surimi stability during frozen storage (Morsy et al., 2018; Marín et al.,
2018). Across sausages and patties, nanoemulsions have reduced lipid-
oxidation markers by 30–60% and microbial counts by 1–2 log CFU/g
(Pereira et al., 2026). However, instability, volatility, poor solubility
and sensory effects remain recurrent limitations, and results are dif-
ficult to compare when dose and synthetic benchmarks differ
(Haskaraca and Sıçramaz, 2026).
Babatunde and de Souza 10.3389/fsufs.2026.2008486
Frontiers in Sustainable Food Systems 13 frontiersin.org
Edible oils illustrate why different stability tests can produce
apparently contradictory conclusions. A pomegranate-peel double
nanoemulsion at 300 ppm controlled oxidation in soybean and mus-
tard oils better than the unencapsulated extract and butylated
hydroxytoluene ( Rashid et al., 2022 ). By contrast, an olive-leaf
water-in-oil nanoemulsion reduced peroxide formation in soybean
oil but performed less well in the high-temperature Rancimat test
because phenolics remained confined within droplets (Mohammadi
et al., 2015). The apparent contradiction reflects different oxidation
conditions and access of antioxidants to the reaction interface; no
single accelerated test is sufficient to predict shelf-life
performance.
TABLE 5 Food-matrix performance of selected residue-derived formulations.
Residue and
cargo
Recovery Carrier (size; EE) Food matrix Outcome versus
control
Source
Pomegranate peel
phenolics
Ultrasound-assisted Double W/O/W
nanoemulsion, MD/WPI (108
to 191 nm)
Soybean and mustard
oils
300 ppm outperformed
free extract and BHT in
controlling oxidation
Rashid et al. (2022)
Pomegranate peel
extract
Lyophilised Peel nanoparticles Beef meatballs, 15 d
at 4 °C
Lower peroxide, TBARS
and TVB-N; acceptable
colour and odour
Morsy et al. (2018)
Acerola by-product
phenolics
Extraction in chitosan
solution
Chitosan-TPP nanoparticles Guava (coating) Delayed ripening;
retained green colour and
firmness
Silva et al. (2021)
Orange peel essential
oil
NR Nanoemulsion in pectin
coating
Fresh-cut orange, 17
d at 4 °C
Higher antimicrobial
effect than
microemulsion; no
sensory difference
Radi et al. (2018)
Tomato peel extract Solvent extraction Zein and gelatin electrospun
fibres
Y ogurt Radical scavenging up 40
to 60%; colour shift
İnanç Horuz and
Belibağlı (2019b)
Banana and apple peel
extracts
NR Basil seed gum nanoparticles
(112 to 117 nm)
Whipped cream, 28 d Encapsulated form more
antimicrobial; free form
more antioxidant
Shafahi et al. (2025)
Citrus reticulata peel
β-carotene
Solvent extraction Ultrasonic nanoemulsion
(143.7 nm)
Fruit juice Enhanced colour and
β-carotene bioaccessibility
Barman et al. (2020)
Pomegranate peel and
other waste extracts
NR Soy phosphatidylcholine
liposomes (76 to 81 nm fresh)
Squid surimi gels,
frozen storage
Maintained stability;
slight loss of gel strength
Marín et al. (2018)
Sour cherry pomace
phenolics
Solvent extraction Whey and soy protein
encapsulates
Cookies, 4 months Anthocyanins fell 35.9 to
67.9%; acceptable sensory
quality
Tumbas Šaponjac et al.
(2016)
Pomegranate peel
polyphenols
Solvent extraction Orange juice by-product wall Cookies (5,000 ppm
phenolics)
Degraded in baking but
retained better than free
extract; no sensory
penalty
Kaderides et al. (2019)
Grape, jaboticaba and
hibiscus by-product
phenolics
Green hydroalcoholic
extraction
Chitosan-TPP (546 to
1,539 nm; EE up to 87.5%)†
Tapioca heated to
200 °C
More than 89% phenolic
retention
Oliveira et al. (2026)
Grape pomace
phenolics
Solvent extraction Spray-dried maltodextrin and
gum arabic (3.48 μm; EE
97.8%)†
Greek-style yogurt,
90 d
Stable encapsulate; higher
antioxidant activity and
intestinal bioaccessibility
López-Astorga et al.
(2025)
Kiwifruit peel extract Ultrasound-assisted Encapsulated extract Beef patties, 9 d,
MAP
Better oxidative stability
than erythorbate; no
antimicrobial effect
Sirini et al. (2025)
Olive leaf extract Solvent extraction W/O nanoemulsion (6.16 nm)
and multiple emulsions
Soybean oil Lower peroxide value
than free extract; lower
Rancimat stability
Mohammadi et al.
(2015)
Full table in Supplementary Table S1. All studies were laboratory scale. †Above the <1 μm working definition; retained for food-performance evidence. BHT, butylated hydroxytoluene; EE,
encapsulation efficiency; MAP , modified-atmosphere packaging; MD, maltodextrin; NR, not reported; TBARS, thiobarbituric acid reactive substances; TPP , tripolyphosphate; TVB-N, total
volatile basic nitrogen; W/O, water-in-oil; W/O/W , water-in-oil-in-water; WPI, whey protein isolate.
Babatunde and de Souza 10.3389/fsufs.2026.2008486
Frontiers in Sustainable Food Systems 14 frontiersin.org
Fresh-produce coatings can combine a physical moisture or gas
barrier with controlled release of residue-derived antimicrobials and
antioxidants. Chitosan nanoparticles containing acerola by-product
phenolics delayed guava ripening and preserved colour and firmness,
orange-peel essential-oil nanoemulsions in pectin coatings extended
the life of fresh-cut orange, and apple-pomace lignocellulosic nanofi-
bres formed coatings for fresh-cut apple (Silva et al., 2021; Radi et al.,
2018; Zhang S. et al., 2024). Their practical value depends on uniform
coverage, adhesion, sensory neutrality and migration behaviour,
which are reported less consistently than short-term quality outcomes.
Active and intelligent packaging combines residue extracts with
residue-derived nanofillers to modify barrier, mechanical, antioxidant
or indicator properties. Applications have been demonstrated for
strawberries, beef, chicken and seafood (Ma et al., 2025; Xin et al.,
2026; Kargar et al., 2026; Jiang et al., 2024). Most evidence concerns
film properties, while contact-food tests are short and migration, toxi-
cology and end-of-life behaviour are seldom integrated. Model-film
success therefore cannot yet establish safe and scalable food-packaging
performance.
6.2 Matrix-driven failure modes
Nanocarriers that remain stable in buffer may fail after contact
with food. Protein-stabilised nanoemulsions are sensitive to pH,
ions, heating and freeze-thawing ( Adjonu et al., 2023 ).
Mechanistically, protein interfaces lose electrostatic repulsion as
pH approaches the isoelectric point or as salts screen surface
charge, divalent cations bridge anionic polymers, and surfactants
or bile salts displace interfacial proteins; these processes govern
aggregation and release ( Tan et al., 2019 ). Nanoemulsions are vul -
nerable to Ostwald ripening, flocculation, coalescence, creaming
and phase separation, with the dominant pathway determined by
oil composition, emulsifier, temperature, pH and ionic strength
(Karthik et al., 2017 ; Marhamati et al., 2021 ). Real-food reviews
similarly show that a matrix can reduce bioavailability and anti -
microbial efficacy relative to model systems ( Nejatian et al., 2022 ).
The pH-, salt- and heat-dependent stability of potato-protein
nanoemulsions is particularly relevant to residue-derived pro -
teins, whose composition may be less uniform ( Tan et al., 2022 ).
Figure 4 summarises the principal matrix-induced failure
mechanisms.
• Ionic strength and pH. Increasing salt concentration from 0 to
400 mM enlarged droplets in tea seed cake protein Pickering
emulsions (Liang et al., 2022), and citrus peel extracts degraded
faster at low pH (Hu et al., 2019).
• Aggregation in solid matrices. At higher loadings, lemon peel
extract nanogels accumulated on film surfaces ( Mortazavi
Moghadam et al., 2023), and anthocyanin extract reduced elonga-
tion at break of pullulan-polyvinyl alcohol nanofibre films from
24.7 to 14.9% (Jiang et al., 2024).
• Texture, structure and colour. Liposomes slightly weakened surimi
gels (Marín et al., 2018), nanoencapsulated beetroot peel extract
softened curd (Pandey and Singh, 2025), and encapsulated pig-
ments shift product colour, a defect in yogurt (İnanç Horuz and
Belibağlı, 2019a) but a function in beverages coloured with carot-
enoid nanoemulsions (Barman et al., 2020).
• Binding, astringency and carrier-lipid interactions. Pomegranate
peel ellagitannins bind food proteins and salivary glycoproteins
(Andishmand et al., 2023), interactions of citrus essential oils with
food components remain a central obstacle to their use (Koch et
al., 2024), and the choice of oil changed the performance of olive
and tomato pomace nanoemulsions in a beverage ( Katsouli et
al., 2024).
These mechanisms are seldom measured directly in foods over
realistic storage. Because product responses can improve while
FIGURE 4
Common failure modes of nanocarriers in food matrices: (A) Flocculation and coalescence; (B) Ostwald ripening and creaming; (C) Aggregation in solid
matrices; (D) Cargo-protein binding and astringency; (E) Texture and structure change; (F) Colour drift.
Babatunde and de Souza 10.3389/fsufs.2026.2008486
Frontiers in Sustainable Food Systems 15 frontiersin.org
particles aggregate, and intact carriers may release too little cargo, size,
retention and release should be tracked within the matrix.
6.3 Sensory and consumer acceptance
Encapsulation may improve acceptability by masking undesirable
colour, as reported for djenkol-peel-extract liposomes, or by reducing
astringency, as proposed for cashew-nut-testa phenolics
(Chotphruethipong et al., 2025; Sruthi et al., 2024). Y et sensory benefit
should not be inferred from nanoscale size: panels detected no differ-
ence between microemulsion and nanoemulsion coatings on fresh-cut
orange (Radi et al., 2018). Evidence remains preliminary: panels are
small, dose–response and long-term acceptance are rarely tested, and
willingness to pay has not been evaluated.
6.4 Where benefit is demonstrated and
where claims outrun evidence
The strongest application evidence compares the nanoencapsu -
lated extract with the free extract and an appropriate synthetic
benchmark in a real food over its expected shelf life. Pomegranate-
peel nanoemulsions tested against butylated hydroxytoluene in
edible oils and pomegranate-peel nanoparticles tested in meatballs
illustrate this design (Rashid et al., 2022; Morsy et al., 2018). Bakery
studies that follow storage for months and meat studies that include
a synthetic control provide similarly useful models ( Tumbas
Šaponjac et al., 2016; Sirini et al., 2025 ). Across matrices, the most
consistent benefit is oxidative stability in oils and meat; antimicrobial
effects are less consistent (Sirini et al., 2025 ) and sensory benefit is
rarely shown. The evidence supports this, but not the stronger claim
that nanoencapsulation will outperform free extracts across food
matrices.
7 Safety, regulation and scale
7.1 Nanomaterial safety and gastrointestinal
exposure
Safety assessment must separate hazards associated with the car-
rier from those associated with the residue-derived cargo. Organic
particles made from lipids, proteins or carbohydrates are generally
hydrolysed in the gastrointestinal tract, which reduces the likelihood
of persistent-particle exposure. Inorganic particles can persist, how-
ever, and their fate depends on size, aggregation, surface coating, crys-
tal form and food-matrix interactions (McClements and Xiao, 2017).
Most residue-derived systems use organic carriers, but mesoporous
silica, selenium nanoparticles and zeolitic imidazolate frameworks
introduce less understood gastrointestinal and toxicological questions
(Jokarshourijeh et al., 2025 ; Rasheed, and Jum’a, 2026; Xin et al.,
2026). Small droplets can also alter digestion and absorption without
persisting as particles, so rapid degradation does not remove the need
for exposure-based assessment ( McClements and Rao, 2011 ;
McClements, 2020).
Toxicological evidence is not yet proportionate to the functional
claims made for these systems. Safety assessment is commonly limited
to one cell line, such as hippocampal HT22 cells for guava-seed
nanoparticles or HepG2 cells for beef-packaging films (Schappo et al.,
2026; Xin et al., 2026). Such tests are useful for screening but cannot
establish oral safety, chronic exposure or interactions with digestion.
Assessment should cover both carrier and cargo, transformation prod-
ucts formed during processing or digestion, intended use levels and
the resulting dietary exposure. Natural origin is not evidence of safety,
and a carrier designed to increase bioavailability may also increase
internal exposure to both desired and undesired compounds (Vilas-
Boas et al., 2021; McClements and Xiao, 2017). Food components can
further change nanoparticle properties and nutrient absorption, while
acute oral and longer-term in vivo studies remain uncommon (Cao et
al., 2016; de Oliveira Mallia et al., 2021; Zhang L. et al., 2025; Shukla
et al., 2017; Jasim et al., 2024).
7.2 Regulatory positions
Binding legislation and scientific guidance should be distin -
guished. Under Regulation (EU) 2015/2283, food consisting of
engineered nanomaterials is a novel food requiring authorisation,
and Regulation (EU) No 1169/2011 requires such ingredients to be
labelled “(nano)”; in plastic food-contact materials, substances in
nanoform need specific authorisation under Regulation (EU) No
10/2011 (European Commission, 2011 , 2026). EFSA guidance is not
legislation but sets the data expected, including physicochemical
characterisation and tiered testing of in vitro degradation, toxico -
kinetics and genotoxicity ( Hardy et al., 2018 ; Scientific Committee
et al., 2021). The 2022 Commission Recommendation on the defini -
tion of nanomaterial is also non-binding; a delegated act aligning
the novel-food definition with it was rejected by the European
Parliament in April 2024, so the definition in Regulation (EU)
2015/2283 still applies ( European Commission, 2022 , 2026;
European Parliament, 2024 ).
Scope relative to this review. The review’s < 1 μm working defini-
tion is broader than the EU criterion. Regulation (EU) 2015/2283
refers to dimensions of the order of 100 nm or less, and the 2022
Recommendation applies a number-based 1–100 nm range to at least
50% of particles. Many systems reviewed here, including most in
Table 5, therefore fall outside the engineered-nanomaterial scope on
reported mean size. Mean size alone cannot settle classification, which
also depends on the number-based size distribution, intentional
design, retention of nanoscale properties and the sector-specific
definition.
United States. The Food and Drug Administration does not con-
sider nanotechnology inherently safe or harmful. Its 2014 guidance
states that intentional reduction of particle size to the nanoscale may
affect the identity, safety or regulatory status of a food substance,
including its status as generally recognised as safe, and should be
assessed case by case (FDA, 2014).
African jurisdictions. Evidence is limited and dated: a 2015 com-
parative review found that South Africa regulated nanotechnology-
based food products under general foodstuffs legislation (Amenta et
al., 2015), and national rules should be checked individually. Globally,
regulation is complicated by the lack of a harmonised definition and
by differing regulatory scopes (Allan et al., 2021), and a 2025 system-
atic review of agro-nanotechnology in Africa documents growing
research activity but focuses on agricultural inputs rather than food
ingredients (Gelaye, 2025).
Babatunde and de Souza 10.3389/fsufs.2026.2008486
Frontiers in Sustainable Food Systems 16 frontiersin.org
7.3 Contaminant carry-over from residues
Residues can concentrate contaminants alongside bioactives.
Structured hazard identification across nine plant side streams identi-
fied heavy metals, mycotoxins and plant-protection products as
recurring concerns; citrus peel presented the broadest potential
hazard profile, while beetroot, sugar beet and carrot presented fewer
identified hazards (van Asselt et al., 2025). The mechanism is impor-
tant: citrus processing enriched some pesticide residues in essential
oil by factors of 1.68–39.2, certain sweet-cherry pits exceeded maxi-
mum residue levels for five pesticides, grape seeds contained about
130 mg/kg aluminum together with arsenic, cadmium and lead, and
grape pomace contained 26–100 μg/kg ochratoxin A, which autoclav-
ing reduced by 19–80% (Tang et al., 2023; Mateus et al., 2024; Yu et
al., 2020). Because hydrophobic pesticides and mycotoxins may parti-
tion with lipophilic cargo, a process optimised for recovery could also
enrich hazards. This plausible carry-over has rarely been measured
in final carriers. Processing can add further hazards: drying or roast-
ing can form polycyclic aromatic hydrocarbons and acrylamide
(Ververis et al., 2026), mycotoxin co-contamination is frequent in
by-products such as grape pomace (Lopes et al., 2023), and protein-
rich residues such as seed cakes can carry allergens into residue-
derived protein carriers (Boopathi et al., 2026). EFSA assessments of
by-product-derived novel foods accordingly require identity, produc-
tion process, composition, intended use and exposure, toxicology and
allergenicity data (Precup et al., 2024). Extracts and finished ingredi-
ents should therefore be screened for pesticides, mycotoxins and
heavy metals using matrix-adapted multi-residue methods and
exposure-based risk assessment (Tadesse Tsegay et al., 2025; Ververis
et al., 2026).
7.4 Techno-economics and scale-up
Commercial feasibility depends less on the nominally low cost of
residues than on process yield, energy demand, solvent and water
recovery, drying, equipment utilisation and consistency of supply
(Shuaibu and Juma, 2026). Spray drying is the most established con-
tinuous route, and models can predict moisture, solids recovery and
phenolic retention for grape-pomace and citrus-peel extracts
(Saborirad et al., 2024; Assadpour and Jafari, 2019; Jayaprakash et al.,
2022; Bassani et al., 2022). Freeze-dried nanoemulsion powders and
proniosomes offer storable formats, but freeze-drying is energy inten-
sive and can enlarge particles; electrohydrodynamic routes operate
under mild conditions but remain less established at industrial scale
(Rashid et al., 2022; Ilgaz et al., 2025; Marín et al., 2018; Jacobsen et
al., 2018; Tripathy et al., 2026). Residue seasonality and compositional
variability add supply risk that must be managed by stabilisation,
blending or extract specifications. Where assessed, energy dominates
both cost and impact: for freeze-dried pomegranate-peel extract
microcapsules, freeze-drying contributed 75–77% of environmental
burden and production cost fell to about US$7 per kg only at 5 tonnes
per day (Fajimi et al., 2026), and energy optimisation governed both
profitability and impact for orange-peel and tomato-seed carotenoids
(Arias et al., 2025 ). Economic feasibility
and environmental benefit are distinct and can diverge (Okoro et al.,
2023). Laboratory success therefore demonstrates technical possibility,
not commercial feasibility; TEA and LCA remain missing for nano-
encapsulated residue ingredients (Ingale et al., 2025).
8 Integrated outlook
8.1 The recovery-to-application chain as a
single design problem
Each stage changes the material presented to the next (Figure 1),
which explains why performance rarely transfers between studies.
Robust design must therefore be optimised from residue specification
to final-food function, rather than as separate extraction and encap-
sulation experiments.
8.2 Priority research gaps
The most urgent needs are standardised head-to-head compari-
sons, marker-specific metrics, real-food validation, safety assessment
and integrated TEA and LCA (items 2–4, 6 and 7).
(1) Variable inputs. Report a full chemical specification of the
crude extract before formulation (marker compounds, co-
extractives, residual solvent and contaminants), test cultivar,
maturity, stabilisation and recovery route as variables, and
assess batch-to-batch robustness of carriers.
(2) Head-to-head comparisons. Compare carriers with the same
extract, the same markers and the same stress tests, including
residue-derived carriers against conventional wall materials.
(3) Metrics beyond encapsulation efficiency. Report payload, reten-
tion under food-relevant stress and delivered dose alongside
EE (Table 4).
(4) Real foods and realistic storage. Test carriers in the target food
over its expected shelf life, with free-extract and synthetic con-
trols, and track carrier integrity in the matrix.
(5) Digestion and human relevance. Use harmonised digestion
models such as INFOGEST, report oil, emulsifier and calcium
levels, state model limitations (Lu et al., 2020; Brodkorb et al.,
2019; Duque-Soto et al., 2022), and move promising systems
towards in vivo and human studies (Niu et al., 2020).
(6) Safety and contaminants. Screen residue ingredients for pesti-
cides, mycotoxins and heavy metals using structured hazard
lists for each side stream (van Asselt et al., 2025), and apply the
EFSA tiered framework to inorganic and hybrid carriers.
(7) Scale and sustainability. Couple laboratory optimisation with
process models (Bassani et al., 2022), techno-economic and
life-cycle assessment, especially for energy-intensive steps
such as freeze-drying and for NADES-based routes (Gómez-
Urios et al., 2026).
(8) Coverage and transferability. Extend research beyond well-
characterised peel systems to pomaces, seeds, husks, leaves
and mixed industrial streams. Greater attention is also needed
for residues, diets, manufacturing conditions and regulatory
contexts in Africa and other under-represented regions. The
aim should be to test whether design principles transfer across
materially different streams, not simply to expand the number
of residue examples.
9 Conclusion
Nanoencapsulation can stabilise and deliver residue-derived bioac-
tives, but performance is set along the whole chain from residue to
Babatunde and de Souza 10.3389/fsufs.2026.2008486
Frontiers in Sustainable Food Systems 17 frontiersin.org
digestion, not by the carrier alone. Residue variability and recovery
chemistry determine what reaches the carrier, and EE or particle size
alone does not predict food or nutritional benefit. Future studies should
report a complete chemo-analytical specification of the crude extract
before formulation, use harmonised metrics and the INFOGEST proto-
col, test carriers in real foods with free-extract and synthetic controls, and
screen for contaminants. Circularity and sustainability claims require
mass balance, TEA and LCA. Meeting these requirements would move
carrier selection from case-by-case optimisation to transferable design.
Author contributions
JB: Conceptualization, Data curation, Formal analysis, Writing –
original draft. PS: Conceptualization, Supervision, Writing – review
& editing.
Funding
The author(s) declared that financial support was not received for
this work and/or its publication.
Acknowledgments
The authors acknowledge Mohammed VI Polytechnic University
(UM6P) and the African Sustainable Agriculture Research Institute
(ASARI) for the institutional support provided during the preparation
of this work.
Conflict of interest
The author(s) declared that this work was conducted in the
absence of any commercial or financial relationships that could be
construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that Generative AI was used in the
creation of this manuscript. During the preparation of this work,
the authors used ChatGPT (OpenAI, GPT-5.6 Sol model, medium
reasoning mode; https://chatgpt.com ) for language refinement,
manuscript organization, and the generation of conceptual illus -
trations for Figures 1–4. The authors defined the scientific content
and prompts, reviewed and revised the generated outputs, and
verified all labels, relationships, and interpretations against the
evidence presented in the manuscript. No experimental data or
research results were generated, modified, or interpreted by
AI. All scientific interpretations, literature selection, evidence
assessment, and conclusions were determined and verified by the
authors. The authors take full responsibility for the accuracy,
originality, and integrity of the manuscript and figures.
Any alternative text (alt text) provided alongside figures in this
article has been generated by Frontiers with the support of artificial
intelligence and reasonable efforts have been made to ensure accuracy,
including review by the authors wherever possible. If you identify any
issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the
authors and do not necessarily represent those of their affiliated
organizations, or those of the publisher, the editors and the
reviewers. Any product that may be evaluated in this article, or
claim that may be made by its manufacturer, is not guaranteed or
endorsed by the publisher.
Supplementary material
The Supplementary material for this article can be found online
at: https://www.frontiersin.org/articles/10.3389/fsufs.2026.2008486/
full#supplementary-material
References
Adjonu, R., Doran, G., Torley, P ., and Agboola, S. (2023). Stability of whey protein bioac-
tive peptide-stabilised nanoemulsions: effect of pH, ions, heating and freeze-thawing. Int.
J. Food Sci. Technol. 58, 1787–1794. doi: 10.1111/ijfs.16292
Ai, C., He, B., Luo, X., Wang, R., Kang, J., Li, X., et al. (2026). Fabrication of a quaternary
ammonium chitosan/pomelo peel extract nanocarrier system for tea polyphenol delivery:
characterization, stability, digestibility, and release properties. Food Chem. X 37:104114.
doi: 10.1016/j.fochx.2026.104114
Akcicek, A., Bozkurt, F ., Akgül, C., and Karasu, S. (2021). Encapsulation of olive pomace
extract in rocket seed gum and chia seed gum nanoparticles: characterization, antioxidant
activity and oxidative stability. Foods 10:1735. doi: 10.3390/foods10081735
Akram, A., Murugan, J. S., Barathikannan, K., Tammina, S. K., Sridhar, K., Keewan, M.,
et al. (2025). Pea protein-potato peel starch complex coacervation for the encapsulation
of bioactive date seed extract: bio-functional characterization and in vitro release. Int. J.
Biol. Macromol. 310:143484. doi: 10.1016/j.ijbiomac.2025.143484
Alañón, M. E., Pimentel-Moral, S., Arráez-Román, D., and Segura-Carretero, A. (2021).
HPLC-DAD-Q-ToF-MS profiling of phenolic compounds from mango (Mangifera indica
L.) seed kernel of different cultivars and maturation stages as a preliminary approach to
determine functional and nutraceutical value. Food Chem. 337:127764. doi: 10.1016/j.
foodchem.2020.127764
Aliakbarian, B., Paini, M., Adami, R., Perego, P ., and Reverchon, E. (2017). Use of super-
critical assisted atomization to produce nanoparticles from olive pomace extract. Innov.
Food Sci. Emerg. Technol. 40, 2–9. doi: 10.1016/j.ifset.2016.09.016
Allan, J., Belz, S., Hoeveler, A., Hugas, M., Okuda, H., Patri, A., et al. (2021). Regulatory
landscape of nanotechnology and nanoplastics from a global perspective. Regul. Toxicol.
Pharmacol. 122:104885. doi: 10.1016/j.yrtph.2021.104885
Amaya-Cruz, D. M., Pérez-Ramírez, I. F ., Delgado-García, J., Mondragón-Jacobo, C.,
Dector-Espinoza, A., and Reynoso-Camacho, R. (2019). An integral profile of bioactive
compounds and functional properties of prickly pear (Opuntia ficus indica L.) peel with
different tonalities. Food Chem. 278, 568–578. doi: 10.1016/j.foodchem.2018.11.031
Amenta, V ., Aschberger, K., Arena, M., Bouwmeester, H., Botelho Moniz, F ., Brandhoff, P .,
et al. (2015). Regulatory aspects of nanotechnology in the Agri/feed/food sector in EU and
non-EU countries. Regul. Toxicol. Pharmacol. 73, 463–476. doi: 10.1016/j.yrtph.2015.06.016
Andishmand, H., Azadmard-damirchi, S., Hamishekar, H., Torbati, M., Kharazmi, M. S.,
Savage, G. P ., et al. (2023). Nano-delivery systems for encapsulation of phenolic compounds
from pomegranate peel. Adv. Colloid Interf. Sci. 311:102833. doi: 10.1016/j.cis.2022.102833
Andrade, R. C., Ribeiro, C. D. F ., Assunção, L. S., de Souza Caetano, V . C., de Carvalho
Coelho, L. M., Treichel, H., et al. (2026). Red pitaya peel as encapsulant for hybrid crude
palm oil nanoparticles: characterization and 60-day stability. Food Chem. 520:149786. doi:
10.1016/j.foodchem.2026.149786
Arias, A., Feijoo, G., and Moreira, M. T. (2025). Agri-food waste to phenolic compounds:
life cycle and eco-efficiency assessments. J. Environ. Manag. 380:124935. doi: 10.1016/j.
jenvman.2025.124935
Arroyo-Esquivel, L., Rincón, E., Jiménez, V . M., Vásquez, F ., Esquivel, P ., Espinosa, E.,
et al. (2025). Valorization of red pitahaya (Hylocereus sp.) peel through a multi-product
Babatunde and de Souza 10.3389/fsufs.2026.2008486
Frontiers in Sustainable Food Systems 18 frontiersin.org
cascade biorefinery approach towards bioactive food packaging films. Food Biophys.
20:24. doi: 10.1007/s11483-024-09907-w
Assadpour, E. (2018). A systematic review on nanoencapsulation of food bioactive ingre-
dients and nutraceuticals by various nanocarriers. Crit. Rev. Food Sci. Nutr. 59, 3129–3151.
doi: 10.1080/10408398.2018.1484687
Assadpour, E., and Jafari, S. M. (2019). Advances in Spray-Drying Encapsulation of Food
Bioactive Ingredients: from Microcapsules to Nanocapsules. Annu. Rev. Food Sci. Technol.
10, 103–131. doi: 10.1146/annurev-food-032818-121641
Ayoub, M., de Camargo, A. C., and Shahidi, F . (2016). Antioxidants and bioactivities of
free, esterified and insoluble-bound phenolics from berry seed meals. Food Chem. 197,
221–232. doi: 10.1016/j.foodchem.2015.10.107
Baljak, J., Dekanski, D., Pirković, A., Mitić, N., Rašković, A., Kladar, N., et al. (2025).
Valorization of olive leaf extract via tailored liposomal carriers: comparative analysis of
physicochemical features, antioxidant capacity, and stability. Pharmaceuticals 18:1639.
doi: 10.3390/ph18111639
Barman, K., Chowdhury, D., and Baruah, P . K. (2020). Development of β-carotene loaded
nanoemulsion using the industrial waste of orange (Citrus reticulate) peel to improve in
vitro bioaccessibility of carotenoids and use as natural food colorant. J. Food Process.
Preserv. 44:e14429. doi: 10.1111/jfpp.14429
Bassan, L. T., Nascimento, K. R., Choquetico Iquiapaza, I. Y ., da Silva Ferreira, M. E.,
Tapia-Blacido, D. R., Fabi, J. P ., et al. (2025). Chitosan suspension enriched with phenolics
extracted from pineapple by-products as bioactive coating for liposomes: physicochemi-
cal properties and in vitro cytotoxicity. Food Res. Int. 201:115571. doi: 10.1016/j.
foodres.2024.115571
Bassani, A., Carullo, D., Rossi, F ., Fiorentini, C., Garrido, G. D., Reklaitis, G. V ., et al.
(2022). Modeling of a spray-drying process for the encapsulation of high-added value
extracts from food by-products. Comput. Chem. Eng. 161:107772. doi: 10.1016/j.
compchemeng.2022.107772
Boopathi, V ., Kwon, J., and Jung, J. (2026). Agricultural by-products as secondary risk
matrices: Hazard occurrence, contaminant fate, mitigation strategies, and regulatory con-
siderations in food systems. Compr. Rev. Food Sci. Food Saf. 25:e70632. doi:
10.1111/1541-4337.70632
Brodkorb, A., Egger, L., Alminger, M., Alvito, P ., Assunção, R., Ballance, S., et al. (2019).
INFOGEST static in vitro simulation of gastrointestinal food digestion. Nat. Protoc. 14,
991–1014. doi: 10.1038/s41596-018-0119-1
Bruna, L., Marty, C., Draye, M., Cravotto, G., and Chatel, G. (2025). Subcritical or
supercritical? A comparative life cycle assessment of bioactive compound extraction
from apple pomace. Cleaner Environ. Syst. 18:100311. doi: 10.1016/j.
cesys.2025.100311
Cabezas-Terán, K., Grootaert, C., Van Camp, J., Ortiz, J., Ruales, J., Donoso, S., et al.
(2025). Bioaccessibility of β-carotene during in vitro co-digestion of encapsulated mango
peel carotenoids with milk. Food Res. Int. 214:116576. doi: 10.1016/j.foodres.2025.116576
Cai, T., Xiao, P ., Yu, N., Zhou, Y ., Mao, J., Peng, H., et al. (2020). A novel pectin from
Akebia trifoliata var. australis fruit peel and its use as a wall-material to coat curcumin-
loaded zein nanoparticle. Int. J. Biol. Macromol. 152, 40–49. doi: 10.1016/j.
ijbiomac.2020.02.234
Cannavacciuolo, C., Pagliari, S., Frigerio, J., Giustra, C. M., Labra, M., and Campone, L.
(2022). Natural deep eutectic solvents (NADESs) combined with sustainable extraction
techniques: a review of the green chemistry approach in food analysis. Foods. 12:56. doi:
10.3390/foods12010056
Cao, Y ., Li, J., Liu, F ., Li, X., Jiang, Q., Cheng, S., et al. (2016). Consideration of interaction
between nanoparticles and food components for the safety assessment of nanoparticles
following oral exposure: a review. Environ. Toxicol. Pharmacol. 46, 206–210. doi:
10.1016/j.etap.2016.07.023
Carpentieri, S., Ferrari, G., and Donsì, F . (2023). All-natural wheat gliadin-gum arabic
nanocarriers for encapsulation and delivery of grape by-products phenolics obtained
through different extraction procedures. Food Chem. 424:136385. doi: 10.1016/j.
foodchem.2023.136385
Carpentieri, S., Ghanem, A., Khwaldia, K., Silva, A. S., and Ferrari, G. (2025). Life
cycle assessment of agro-industrial residues valorization processes to obtain pheno -
lic-rich extracts. Front. Sustain. Food Syst. 9:1693181. doi: 10.3389/
fsufs.2025.1693181
Chanioti, S., Katsouli, M., and Tzia, C. (2021). Novel processes for the extraction of phe-
nolic compounds from olive pomace and their protection by encapsulation. Molecules
26:1781. doi: 10.3390/molecules26061781
Chedea, V . S., Tomoiagă, L. L., Ropota, M., Marc, G., Ranga, F ., Muntean, M. D., et al.
(2025). Transylvanian grape pomaces as sustainable sources of antioxidant phenolics and
fatty acids: a study of white and red cultivars. Antioxidants 14:1152. doi: 10.3390/
antiox14101152
Chellamboli, C., Muthamizhi, K., Ramalakshmi, S., Karthikeyan, C., Lo, S. C., and
Rajkumar, S. (2025). Recent advances in Agri-food waste valorisation and nanoencapsula-
tion of bioactive compounds for sustainable development. Discov. Food 5:355. doi:
10.1007/s44187-025-00613-1
Chen, Z., Zhong, J., Wang, S., Huang, Z., Zhang, Q., and Zhao, Q. (2025). Rice bran, wheat
bran, and soybean residue-derived cellulose nanofibers for stabilizing Pickering emul-
sions and protecting β-carotene. Food Chem. 496:146819. doi: 10.1016/j.
foodchem.2025.146819
Chiang, Y .-C., and Chiang, P .-Y . (2026). Self-emulsifying delivery systems as a valorization
approach for tomato pomace: implications for lycopene bioaccessibility and lipid diges-
tion. Food Biosci. 79:108707. doi: 10.1016/j.fbio.2026.108707
Chotphruethipong, L., Sinthusamran, S., Benjakul, S., Senphan, T., Nalinanon, S., and
Sriket, C. (2025). Development of djenkol peel extract-loaded liposome as functional food
ingredients: physicochemical characteristic, antioxidant activities, and cytotoxicity. J.
Agric. Food Res. 21:102003. doi: 10.1016/j.jafr.2025.102003
de Oliveira Mallia, J., Galea, R., Nag, R., Cummins, E., Gatt, R., and Valdramidis, V .
(2021). Nanoparticle food applications and their toxicity: current trends and needs in risk
assessment strategies. J. Food Prot. 85, 355–372. doi: 10.4315/jfp-21-184
Della Posta, S., Gallo, V ., Gentili, A., and Fanali, C. (2022). Strategies for the recovery of
bioactive molecules from deep eutectic solvents extracts. TrAC Trends Anal. Chem.
157:116798. doi: 10.1016/j.trac.2022.116798
Dellapina, G., Poli, G., Moscatelli, V ., Magalhães, D., Vilas-Boas, A. A., and Pintado, M.
(2025). Flavonoid-rich extracts from lemon and orange by-products: microencapsulation
and application in functional cookies. Foods. 14:3346. doi: 10.3390/foods14193346
Domínguez-Rodríguez, G., Marina, M. L., and Plaza, M. (2022). In vitro assessment of
the bioavailability of bioactive non-extractable polyphenols obtained by pressurized
liquid extraction combined with enzymatic-assisted extraction from sweet cherry (Prunus
avium L.) pomace. Food Chem. 385:132688. doi: 10.1016/j.foodchem.2022.132688
Duque-Soto, C., Quintriqueo-Cid, A., Rueda-Robles, A., Robert, P ., Borrás-Linares, I.,
and Lozano-Sánchez, J. (2022). Evaluation of different advanced approaches to simulation
of dynamic in vitro digestion of polyphenols from different food matrices: a systematic
review. Antioxidants. 12:101. doi: 10.3390/antiox12010101
El-Ashaal, E. S., Elshoky, H. A., El-Sayed, N. M., and Elwahash, N. A. A. (2025). Enhanced
preservation of Domiati cheese using edible nanocoating's of carboxymethyl chitosan
loaded with pomegranate peel extract. Food Res. Int. 219:116993. doi: 10.1016/j.
foodres.2025.116993
Espinosa, E., Rincón, E., Morcillo-Martín, R., Rabasco-Vílchez, L., and Rodríguez, A.
(2022). Orange peel waste biorefinery in multi-component cascade approach: polyphe-
nolic compounds and nanocellulose for food packaging. Ind. Crop. Prod. 187:115413. doi:
10.1016/j.indcrop.2022.115413
Esquivel, P ., Viñas, M., Steingass, C. B., Gruschwitz, M., Guevara, E., Carle, R., et al.
(2020). Coffee (Coffea arabica L.) by-products as a source of carotenoids and phenolic
compounds: evaluation of varieties with different peel color. Front. Sustain. Food Syst.
4:590597. doi: 10.3389/fsufs.2020.590597
European Commission (2011). Commission regulation (EU) no 10/2011 of 14 January
2011 on plastic materials and articles intended to come into contact with food. Official J.
European Union, L 12, 1–89.
European Commission (2022). Commission recommendation of 10 June 2022 on the
definition of nanomaterial (2022/C 229/01). Off. J. Eur. Union C 229, 1–5.
European Commission. (2026). Nanomaterials. Food safety: novel food. Available online
at: https://food.ec.europa.eu/food-safety/novel-food/nanomaterials_en (Accessed
September 16, 2026).
European Parliament. (2024). Resolution of 24 April 2024 on the commission delegated
regulation of 14 march 2024 amending regulation (EU) 2015/2283 as regards the defini-
tion of 'engineered nanomaterials' (C(2024)01612 – 2024/2691(DEA)).
P9_TA(2024)0316.
Fajimi, L. I., Ajala, O. O., S
论文原文与出版信息
Frontiers in Sustainable Food Systems