1. IntroductionStratum corneum (SC) acts as an efficient barrier against physical, chemical, and microbiological xenophobes, preventing their penetration into the skin. However, this excellent barrier is a limiting factor for the penetration of cosmetic active ingredients (also known as actives) into the skin. Skin delivery from topical formulations is known to be very inefficient, with typical bioavailability of less than 2% of the applied dose [1]. A good example is caffeine, a well-studied cosmetic and pharmaceutical active ingredient, also a model hydrophilic compound in skin toxicology. Summarising a series of studies conducted with different topical caffeine formulations, a review article [2] has established that the highest penetration from conventional ointment formulations was only 0.0062%.

Therefore, it is crucial to explore all available means for more efficient delivery of topical (cosmetic and pharmaceutical) active ingredients into the skin. Many technologies have been studied and developed so far, including penetration enhancers, supersaturation, and a wide range of skin delivery systems (e.g., liposomes, niosomes, transfersomes, lipid nanoparticles, polymeric microparticles and nanoparticles, patches, and microneedles). One of the relatively recent approaches is the use of 3D printed platforms (carriers).

3D printing (3DP) is a manufacturing technology that produces 3D objects from a design file using layer-by-layer deposition of material. It offers some advantages over traditional manufacturing techniques, such as one-step fabrication and customisation [3]. In addition, 3D printing has shown potential in increasing skin delivery efficacy and user compliance [4]. The healthcare and medical industry has already benefited from 3DP with versatile applications, from 3D printed pharmaceuticals in solid and semisolid forms [5,6], to those with complex release profiles [7]. In addition, there are 3D printed medical devices, such as patient-specific implants and hydrogel grid wound dressings [8,9], many of them approved by the United States Food and Drug Administration (FDA) [10]. However, the number of applications of 3DP in skin delivery is relatively low, with limited choice of 3DP-specific materials being the biggest obstacle. This is because specific physico-chemical properties, such as photosensitivity or thermal sensitivity, are required for the solidification process of the inks during 3D printing in order to provide the structure of 3D objects; in addition, some 3D printing technologies require the ink to be within certain viscosity range [11]. Another obstacle is high initial investment necessary to increase the production output. Extensive studies in skin delivery have only been carried out in the last two decades [12], and have demonstrated a considerable potential of 3DP in this area. There are already comprehensive review articles which cover different aspects on cosmetic skin delivery platforms [13,14], as well as 3D printed transdermal [3,11,12] and topical [15] drug delivery application. However, none of them discusses cosmetic applications of 3D printed skin delivery platforms in detail. Therefore, it was of interest to summarise the developments in this cutting-edge technology, with the view of its application to the delivery of cosmetic active ingredients. 3. Types of 3D Printed Delivery Platforms

Skin patches and microneedles (MNs) have emerged as the two main types of 3D printed platforms. Due to the same principles of skin delivery of cosmetic and topical pharmaceutical formulations, the developments in both will be reviewed in this section.

Skin patches are the most used and studied among all device-based skin delivery systems. They have a long history for treating skin conditions [42] and have also been used for transdermal delivery. A recent review article reported research work conducted on conventional skin patches, in terms of their active ingredients, materials, delivery enhancers, characterisation methods, and results [43]. Standard methods for the fabrication of skin patches include solvent, hydrogel, and hot melt-based means [44]. Patches closely adhere to the skin and could be designed with or without separate adhesive support, which will be either loaded with active ingredients or saturated with active ingredients from the reservoir [45], as demonstrated in Figure 1. The adhesive property of patches strongly affects the delivery of active ingredients, and in turn their efficacy [46,47]. They create a continuous occlusion which increases skin penetration by providing a strong driving force for the diffusion of active ingredients [48]. Theoretically, cosmetic patches have the potential to tackle many cosmetic skin problems, such as wrinkles, pigmentation, and the effects of aging [49]. The main disadvantage of conventional patches is the low quantity of active ingredients that could be loaded and delivered. The mechanism of drug absorption from a patch-like device starts with its release from the patch, its penetration and then storage in the stratum corneum, at which point it might crystalise and prevent further transport [50]. The next stage is diffusion of the active to the deeper layers of skin, and, if applicable, into the systemic circulation, causing a controlled delay of the therapeutic effect [50]. Another disadvantage is a possible inconsistent diffusion rate of the active ingredient from skin patches, which depends on the skin condition of individuals. This includes skin hydration state, age, and ethnicity, with the SC being a rate-limiting barrier. In addition, the delivery strongly depends on the type and physicochemical properties of the active [3]. Conventional adhesive patches have multilayer structures and are classified by the layer in which the drug/active has been loaded [46,47,51,52], as shown in Figure 1. Therefore, the fabrication involves a multi-stage process [53]. In contrast, 3D printed patches are mostly made of one material in a single layer, although there is a potential for the multi-layered printing. Due to the flexible nature of the 3D printed patches, researchers have been exploring the potential of drug-loaded 3DP mesh or 3DP grid patches for implantation within tissues and to support organs [54].

Overall, the development of 3DP patch platforms is an ongoing research area, with a potential to improve current skin delivery designs.

Microneedles (MNs) have evolved as a hybrid of two conventional skin delivery systems, skin patches and hypodermic injection needles, with some advantages of both, such as ease of administration, being minimally invasive, and enabling high bioavailability of active ingredients [55]. An MN platform is typically composed of micro-sized, needle-like structures attached to a patch (backing) for ease of application. This delivery platform is sometimes referred to as a MNP (microneedle patch). In comparison to patches and semisolid topical formulations, microneedles are more versatile and more efficient in delivering active ingredients into deeper layers of the skin. Instead of relying on passive diffusion, microneedles can actively enhance the delivery by piercing the epidermis (0.1–0.2 mm) and, if so designed, the dermis (1–2 mm), creating microscopic channels [56]. These microscopic punctures could overcome the skin barrier and provide an alternative route for enhanced transport of drug or cosmetic active in a painless, minimally invasive manner [57]. Table 2 provides a summary of the common types of MNs. Some novel MNs not only serve as a delivery platform, but also as a wearable therapeutic device for real time monitoring [58].A recent review paper [13] has provided a comprehensive report on the studies on cosmetic application of microneedles. It summarises MN materials and categorises MN studies in terms of the targeted skin problems.

3D printing could be used in three different ways in the manufacture of an MN platform: (1) to develop ‘male’ master moulds; (2) to coat active ingredients onto previously prepared MNs and (3) to print complete MN structures.

Research on 3D printed cosmetic microneedles is still in its early age. However, several published papers and patents have shown the feasibility of delivering both hydrophilic and lipophilic active ingredients by microneedles that are fabricated by methods other than 3DP [63,64,65], proving that this concept is viable. One study has compared the wrinkle improvement by two different delivery platforms (dissolving MNs and standard formulation, both containing hyaluronic acid); after eight weeks of treatment, the MNs have shown higher effectiveness [66].In general, all MNs enhance skin delivery via micro-channels they create, partly bypassing the skin barrier. In the case of wrinkle improvement, there is a second mechanism [67,68]: the perforations they create could induce elastin and collagen expression and deposition, stimulating the metabolism in the upper skin layers, as well as the natural healing of the skin. Table 3 illustrates the mechanism of skin delivery of different types of 3D printed solid MNs, including coated MNs, dissolving/swellable MNs (DMNs), and hollow MNs. Further details and diagrams could be found in several review articles which focus on microneedle skin delivery platforms, including their characteristics and their typical delivery mechanisms [61,69,70]. There is an argument that dissolvable MNs (DMNs) might not be an ideal platform for cosmetic use, due to the potential loss of hydration through the perforations made in the skin. However, it has been shown that a novel design of DMNs, loaded with a barrier-restoring active ingredient, horse oil, has significantly improved dermal density, skin elasticity and moisturisation level [71].Very few published studies have reported a detailed 3DP manufacture of the dissolving MNs, and none has studied the use of sugar-based biopolymers [72]. It remains challenging to fabricate DMNs with 3DP other than by micro-moulding methods, due to the lack of the printability of dissolvable polymers. Fabrication MethodsConventional MN fabrication methods can be classified into several categories: (1) moulding method, (2) lithography, (3) droplet-born air blowing (DAB) method, some followed by coating or deposition process to produce coated MNs [55,70,73,74,75].As described by Kim et al., DAB was a popular DMN fabrication method [76], adopted by many researchers who have successfully fabricated DMNs and assessed their effects in improving skin delivery of cosmetic active ingredients [77,78,79,80]. This method was gradually abandoned with the development of centrifugal lithography (CL) [81] for DMN fabrication. Due to the self-shaping nature of viscous polymer solution, continuous transformation under centrifugal force is induced in the CL process. Polymer drops dispense on the inner plate of two parallel fixed plates, and, upon the separation of plates, DMNs with two different shapes have been formed on the top and bottom plates, respectively. Morphological observation, fracture force analysis, and in vitro skin penetration tests have shown that both DMNs platforms could achieve an efficient diffusion and permeation of active ingredients through the skin [82]. It is worth mentioning that no additional environmental stimulation is required for producing DMNs using CL. The usual problems related to other fabrication methods, such as the loss of activity of cosmetic ingredients when exposed to UV irradiation, heat, and air, do not exist in CL. However, CL-produced MN shapes are extremely limited, with little variation of the natural droplet shape, which points to the necessity of studying the use of 3DP technology in the fabrication of DMNs.

Apart from the use of FDM and SLA, the two most common 3DP technologies, microneedles have also been successfully produced using some novel 3DP technologies, including DLP, CLIP, and TPP.

An investigation on the use of high precision DLP for the 3D printing of hydrogel MNs in terms of the process parameters were performed by Yao et al. [18]. A dye rhodamine B was used as the model compound for the platform characterisation. Its loading was achieved through soaking of the DLP printed MN in the dye solution. The authors have concluded that the long exposure time enhances the stiffness of MNs, and that with the use of hydrogel, the drug loading capacity was greatly increased. There was also a significant decrease in the fabrication time, which only took a few minutes [18].The DLP printing of personalised and flexible MN patches has been extensively studied by Lim et al. [83], e.g., the MN patches to treat the trigger finger, which is not achievable with conventional MNs [83]. Their more recent studies on printing MN periorbital patch focused on the relationship between geometries of these microneedles to their mechanical strength and skin penetration efficiency [84,85]. The fabrication involves two steps: the DLP printing of flat MN patches, which are then compressed against a FDM-printed curved substrate to generate flexible, curved MN patches. Acetyl-hexapeptide 3 (AHP-3) is a small peptide and anti-aging active, with very poor skin penetration due to its hydrophilicity and high molecular weight. With the aid of the optimised DLP-printed MN periorbital patch, enhanced anti-wrinkle effect was achieved with significantly improved AHP-3 delivery [84,85].Different materials and geometries with various aspect ratios of MNs were attempted by Johnson and co-workers [37], using CLIP technology. Square pyramidal needle shape was found to be the most suitable design for encapsulating and delivering a wide range of active ingredients. That shape has been shown to effectively pierce the skin and achieve controlled release of drug, based on the fabrication process of less than 10 min [86]. Another more recent work successfully demonstrated a rapid fabrication method of DIP coating CLIP microneedles for transdermal delivery of therapeutic proteins, achieving a high degree of control over microneedle design parameters [37]. The recent study by Cordeiro et al. [19] has shown that highly precise and reproducible MNs could be successfully manufactured using TPP technology to make silicone MN moulds. MNs with various needle shapes and lengths were then produced by a micro-moulding method. Polyvinylpyrrolidone (PVP) and polyvinyl alcohol (PVA) liquid blend and Polyvinyl methyl ether/maleic acid (PVM/MA) copolymer and poly(ethylene glycol) (PEG) liquid blend were prepared and poured in the silicone moulds to produce dissolvable and ‘super-swellable’ MNs, respectively [19].

The above 3DP technologies generally have high printing resolution and precision, producing micro-sized needle shapes highly suitable for MN-type skin delivery systems. Although significant development in this area has been made, the research is still not widely carried out due to the cost and the need of specialist equipment. It is envisaged that more extensive research will be carried out with further development of 3DP technologies.

4. Materials Used in 3DP PlatformsThis section reviews the materials used in the fabrication of 3D printed skin patches and microneedles, focusing specifically on recent sustainability trends by introducing biopolymers [87]. Common 3DP ink materials adopted for cosmetic-relevant application are listed in Table 4.

Skin patches consist of the backbone support, normally made of polymers, into which the actives could be embedded in different ways.

In their pioneering study at the time, Goyanes et al. [27] used both FDM and SLA methods in the production of the cosmetic patch with anti-acne effect, containing salicylic acid. Several FDM printable soft polymers suitable for cosmetic use were tested, such as polycaprolactone (PCL), thermoplastic polyurethane (NinjaFlex®), and flexible polylactic acid (Flex EcoPLA™). In terms of the loading method, hot melt extrusion was used for the FDM route to dissolve polymer solution in organic solvent with salicylic acid at the concentration of 2% w/w. For the SLA route, the solutions were made up of two photo-linkable materials, poly(ethylene glycol) diacrylate (PEGDA) and poly(ethylene glycol) (PEG) as the photoinitiators, before salicylic acid was added. The results showed that the FDM method was not the right choice; a degradation of salicylic acid occurred due to the high extrusion temperature, while the lower extrusion temperature led to the nozzle blockage. The incorporation of drug also reduced the heat stability of the printable polymers. In contrast, the patches fabricated using SLA method were flexible and smooth, with a higher amount of salicylic acid than that obtained with FDM [27]. A two-layer (PLA/phycocyanin–alginate) novel cosmetic patch was recently studied [115], using a simple casting method for patch fabrication. PLA has been extensively used in the FDM technology, so it was of interest to further study the printability of innovative PLA composite materials. The loaded active, spirulina extract (phycocyanin) has been proven to be safe for cosmetic use as an antioxidant. An optimised formulation has been shown to be a promising patch material, with excellent flexibility and the ability to fully release the active from the phycocyanin–alginate layer of the patch [115]. In the study by Rees et al. on sustainable 3DP materials [116], nanocellulose obtained from renewable sources was used to print a patch-like hydrogel matrix for wound healing. The patch was shown to be effective in inhibiting the growth of common wound pathogen, Pseudomonas aeruginosa. Hence, this platform could be used to tailor-made wound dressings with antimicrobial components. In addition, the porous structure of the 3D printed hydrogel matrix was highly suitable for the incorporation of topical actives, including those for cosmetic applications [116]. A study by Domínguez-Robles et al. [117] has produced a new material by combining naturally occurring polymer lignin with poly lactic acid (PLA). They have used modified hot melt extrusion and FDM technologies for the 3D printing of hydrogel patches for wound dressing. The novel material was proven to be very suitable for the fabrication of 3DP wound dressings, being flexible and able to carry high drug load. Since lignin is abundant, inexpensive, biocompatible, and possesses antioxidant properties, its mixture with PLA might have a wider potential, e.g., for the applications in food packaging, tissue culture, and implants [117]. One difficulty with the traditional cosmetic platform is the loading of a hydrophobic active into a hydrophilic backbone polymer. Sommer et al. [91] have partly solved this problem by 3D printing via direct ink writing (DIW), using an emulsion containing biocompatible soft material. Their oil-in-water emulsion, stabilised by chitosan-modified nano-silica particle, could host both hydrophilic and hydrophobic compounds. Moreover, by printing an active-loaded emulsion onto a regular ink emulsion, the authors achieved a precise positioning of actives within the 3D structure [90]. Since emulsion-based patch is soft and fits the undulated skin, this work opens a new approach for printing cosmetic patches.In line with sustainability efforts, a recent work studied the effect of the incorporation of starch into alginate-based 3D hydrogel patches for topical delivery. The addition of starch increases pore sizes of the matrix after crosslinking, as well as the structural stability. In principle, the drug release rate from 3DP patches is dependent on drug solubility, the influx of the medium into the structure, and polymer(s) swelling. In this case, the in vitro release results showed that release of actives (chemical dye rhodamine B as a model hydrophobic compound) from the starch-incorporated patch was faster and 20% higher in comparison with the patch that did not contain starch [118]. From the sustainability point of view, it is preferable to use biopolymers in 3DP due to their biodegradability under environmental conditions. In addition, most of them provide flexibility and viscoelasticity to the printed patches, while some exhibit antibacterial effect. A major drawback of this type of 3DP patches is their relatively low rate of drug/active release, which is very difficult to improve [3].Microneedles must have sufficient strength to penetrate skin without breakage under the force of thumb [119]. The effectiveness of 3D printed microneedles is mostly determined by the material they are made of. A large variety of materials have been investigated for this purpose, including sugarcane (carboxymethylcellulose-based polymers) [120], glass, metal, carbon-based, and silicon-based polymers [11]. The materials so far studied for the fabrication of the active ingredient-loaded 3DP microneedles include solid polymers (e.g., PLA), hydrogels (e.g., hyaluronic acid), and dissolving polymers (e.g., polyvinyl alcohol) [51,69]. A list of materials that were used for the direct 3D printing of microneedles has been summarised in the work by Economidou and Douroumis [121]. The development of 3DP microneedles is still constrained, since the manufacturing process involves heating, UV, laser light or other instability-causing factors, and requires different viscosities for different 3DP technologies [70]. Selected materials must meet the requirements of non-toxicity and high biocompatibility, while being biodegradable and sustainably produced. Therefore, the development of novel 3D printable material is vital for the development of 3D printed microneedles.In order to penetrate the skin, microneedles have to withstand a certain amount of force without breakage. One of the main concerns about 3D printed polymeric microneedles is their relatively low mechanical strength in comparison with conventional method of moulding. However, it has been shown that successful 3DP microneedles could have comparable mechanical strength to MNs that are constructed by conventional materials [122]. For example, it has been shown that microneedles produced using PLA filament with FDM technology have similar mechanical properties as conventional PLA microneedles made from multistage-micromoulding method [31,123].The investigation by Tang et al. [124] explored the relationship between printing parameters and the quality of the 3D printed transdermal platform in terms of its surface finish and dimensional accuracy, using PLA as the printing material. The resolution and accuracy of printed microneedles were higher when selecting a thinner layer, a smaller infill width and an increasing spacing between needles, while a better surface finish could be achieved using smaller nozzle orifice [124]. The result also revealed that the process of HME and FDM caused temperature-induced degradation of the polymers.Attempts have been made to alter the 3D printable materials in order to overcome the constraints of the technology. Sommer et al. [90] demonstrated a new approach to 3D printing of an oil-in-water emulsion stabilised with chitosan-modified silica. They used direct ink writing (DIW), which is a 3DP method that requires mild conditions. The only prerequisite is that the 3DP ink mixture must have a suitable viscoelasticity, which was achieved through optimisation of the stabiliser, followed by centrifugation to obtain an emulsion paste which would be used as the 3DP ink. The rheological study of the paste showed good alignment with the requirements for the 3DP material and revealed the mixture to be shear-thinning and elastic, with fast recovery [90]. Due to its excellent biocompatibility, viscoelasticity, skin moisturising effect [77] and non-immunogenicity, hyaluronic acid (HA) is the material of choice for the cosmetic dissolving microneedles (DMNs), either on its own or as the backbone material onto which to load other skin actives [125,126]. Recent work by Jang et al. [127] has shown that the active delivery achieved by the HA-based DMNs directly depends on the molecular weight of HA. With the same MN size, geometry and active load, high molecular weight HA performs better than low molecular weight, due to a higher degree of swelling when inserted into the skin. Another use of HA-based DMN was in anti-aging as a skin filler, via the mechanism of short-term epidermal expansion after insertion. Zhang et al. [128] have shown that the stability of the DMNs depends on the degree of HA crosslinking using 1,4-butanediol diglycidyl ether. A higher level of crosslinking produced more stable DMNs, which lasted longer in the skin (2–6 days), therefore it was considered a better filler [128]. With micro-moulding method to fabricate DMNs using 3DP, Shim et al. [129] have presented an innovative approach to optimising skin penetration and the dissolution rate of an active. They have evaluated the use of biocompatible, water-soluble polymers, such as carboxymethylcellulose, hyaluronic acid and polyvinylpyrrolidone (PVP), while the actives were ascorbic acid and acetyl hexapeptide-8. It has been shown that increasing the PVP content of the MNs leads to a faster dissolution of the actives in the skin, but also to their compromised skin penetration. Thus, an optimal composition of the printable materials must be established for each active [129]. DLP 3DP technology has been employed for micro-moulding fabrication method by several research groups [130,131]. In the study by El-Sayed et al. [131], a master mould was produced with DLP, followed by polydimethylsiloxane (PDMS) casting to produce a negative mould. The MN array was then prepared using drug-loaded polymer moulding, vacuuming and base-casting with PVA/sucrose, and its removal after drying. Several different shapes of MNs were also fabricated by DLP [131]. These studies present a useful alternative approach in this area.Other base materials used in cosmetic DMNs include hydroxypropyl methylcellulose (HPMC) and a range of water-soluble polymers, including methylvinyl ether and maleic anhydride copolymer (MVA/MA), polyvinylpyrrolidone (PVP) and chitosan, as well as their combinations [132]. In respect of sustainability, some progress has been made in both the development of novel bioderived [133] or biocomposite [134] 3D printable materials and in the process of production of microneedles using 3DP technology. Clearly, abundant and biodegradable biomaterials should be more widely adopted, as well as 3DP technologies that do not require post-production processes. A close collaboration between these two fields of research is necessary in order to make significant developments in sustainability. Among the reviewed MN types, dissolvable MNs have shown the best sustainability potential, since they could achieve the highest drug/cosmetic active loading, have a single-step application with minimal waste, and could be made of sustainable biomaterials [126]. 5. Characterisation of 3DP Platforms

Testing for stability, safety, and efficacy is a fundamental requirement and must be carried out for both cosmetic and pharmaceutical products. In addition, there is a wide range of physico-mechanical characterisation methods, which could help formulators in their development work, enabling predictions of how products will behave during their production, storage, and use. Some of these methods have the potential to reveal the interactions between active materials and the components of the base. This in turn could explain the observed stability issues, the rate and extent of active ingredient release and ultimately the product efficacy.

Due to the novel nature of the 3DP platforms, a possible interaction (or the lack of it) between the carrier and the active is particularly important. Two aspects of 3DP platforms should be considered: characterisation of finished 3D printed products and characterisation for optimisation of printing process (including intermediates, such as printing filaments).

In terms of the final product, researchers normally report basic physical parameters of the 3DP platforms, including their morphology, geometry, density, and mechanical strength (patch stiffness). For optimising the printing process with extrusion-type 3D printers, the most commonly listed properties for filament polymers include molecular weight, which is measured by gel permeation chromatography (GPC) [124], as well as thermal properties and crystallinity, which are measured by differential scanning calorimetry (DSC) and X-ray diffraction, respectively [135]. These methods are used in addition to well-established standard characterisation techniques for active molecules via in vitro/in vivo permeability studies. Regarding the 3DP patches, few characterisation methods have been used. Rheological properties, gel strength and bio-adhesive properties are key to an effective 3DP patch. Texture analysers and rheometers are mostly used to determine the printability of active -loaded ink, especially important for extrusion-based 3DP technologies [90]. In addition, the pH of the 3DP patch must be suitable for topical use and its pH value, when in contact with moisture, should be close to the pH of human skin [29].For 3DP microneedles, the two properties that have drawn most attention are their geometry and their mechanical properties, which can be further divided into insertion force and mechanical strength (failure test), performed by theoretical simulation and/or experiments [74].Various geometries of MNs have been studied to achieve more defined tip of microneedles, which directly relate to the ability to perforate the SC [136]. The most common shape of MNs are cones, with different aspect ratio, height, interspace, tip diameter and base diameter [121]. There is no standard for the best geometry; in addition, the performance of MNs vary depending on the materials and 3D printers used. Therefore, each parameter must be studied in relation to various ink formulations and 3DP parameters in order to be optimised. Pyramid, cross and spear shapes are also studied, obtained by SLA or other 3DP technologies that have higher resolution [26]. Although FDM printers are easy to use, fast and cost-effective, the technique is generally not suitable for printing the fine structures of MNs. The extrusion manner of printing makes it difficult for printed layers to adhere to one another when the printing area is very small (for sharp tips). Therefore, only shapes that gradually change from the bottom to top could be printed by the FDM method.The observation of platform morphology and the measurement of their dimensions have been carried out using optical microscopy [31,137], scanning electron microscopy [34,137] and in vivo imaging techniques [138]. Image analysis is particularly useful, because it visualises the shape and uniformity of the MN array, allowing checking for any defects [132]. MN platforms are normally applied by pressing them into the skin with a thumb, hence MNs must have sufficient mechanical strength to provide efficient delivery of the actives into the skin [139]. The upper surface of the skin experiences viscoelastic deformation while being perforated with an increasing force. There is a minimal force necessary to punctuate the intact skin, which must not exceed the maximum force that an individual micro-sized needle can withstand, otherwise the needle will break or fracture before piercing the skin [132]. Therefore, it is important to consider mechanical properties of MNs when designing MN platforms. A study by Davis et al. [136] first quantified the effect of geometry to the fracture force of MNs. Their theoretical and experimental analysis both led to the same conclusion: the insertion force varies linearly with the interfacial area of the needle tip [136]. It has been proven by many further studies that the smaller the tip diameter, the easier the perforation [34]. However, the tip diameter is limited by the resolution of the 3D printer, particularly for those using FDM technology. A recent study has shown that, by varying the tilted angle of the MN arrays during the SLA printing process, the tip diameter could be significantly changed [140]. Using the printing angle of 45°, the MNs appeared not only sharper but also without defects. However, the optimisation of printing quality and geometry accuracy differs significantly between the 3DP technologies, so it remains challenging to print sophisticated micro-sized needles.The process of insertion of NMs into the skin has been evaluated by several methods. The penetration test using the membrane that mimics human skin was employed to determine the rate of piercing and the rate of needle breakage after the insertion [31]. Another approach used dye solution applied on the surface of the skin sample, before applying and removing MNs, and analysed the coloured holes produced. In the same study, when the insertion speed was kept constant at 0.5 mm/s, the predicted minimum insertion force through a multilayer skin structure obtained through modelling by Finite Element Analysis for each MN was above 0.03 N. This was consistent with their experimental result of 0.069 N and the literature [34]. Texture analyser has also been employed with skin samples to quantify the insertion of the MN platform, by reporting the continuous force and displacement of microneedle arrays fixed on the top of a moving probe [141]. The mechanical strength or fracture point of MNs were measured in various ways. Transversal, axial, and bending forces were exerted on the MN array to determine the point of mechanical failure by mechanical testers; the shear resistance was also measured [105]. It was found that the 3D printed MNs could be refined post-printing via etching (when using FDM) [31] and post-curing (when using SLA) [140].Since transepidermal water loss (TEWL) reflects the integrity of the skin barrier, changes in TEWL have also been used to evaluate the effects of MNs penetration [29,142].Comprehensive evaluation on the physico-mechanical properties of 3D printed platforms is important for their development and optimisation. For FDM-produced 3DP hydrogel patches and dissolving MNs, the addition of actives may significantly change rheological properties of the formulation, leading to a varied mechanical strength of the MNs after solidification process. In such cases, rheological characterisation is being used to evaluate and regulate their viscosity [68,105]. For developing dissolving MNs, it is vital to understand the process of MNs degradation, since the actives are released during this process. SEM provides information on any change in porosity and formation of cracks in the MN structure, while DSC and X-ray diffraction measure the change of crystallinity of the polymer. Since the crystalline region of the MN is where the integrity of the polymer structure was maintained, amorphous regions start to degrade or dissolve first [30]. 6. Release and Skin Delivery of Actives Used in 3DP Platforms

This section presents a review of the release and penetration studies that have been performed on 3DP platforms in order to study them as carriers for pharmaceutical and cosmetic active molecules.

Even though a series of examples of 3D printed patches for wound healing have been discussed, the delivery mechanism is different from the one occurring in cosmetic application, since the application sites normally do not have a functioning skin barrier.

All published studies related to the use of 3D printing for the delivery of cosmetic active ingredients are summarised in Table 5, including potential ones. The use of standard delivery platforms (patches and MNs) is widely studied [13,14,49,106], but very few attempts have been made with 3D printed skin delivery platforms. Some methods normally used for tissue engineering, wound dressing, and food industry might be transferable for cosmetic applications.Skin permeation study is vital to determine the efficacy of 3DP delivery platform by providing information on any enhancement of the active ingredient delivery in comparison to other delivery systems [70,146]. Ideally, the performance of loaded 3DP platforms should be directly compared to non-loaded controls, or topical formulations containing the same active ingredient. Skin permeation study, which could be performed in vivo or in vitro, is usually preceded by in vitro release study. Franz diffusion cells are the most widely used piece of equipment for the study of drug/cosmetic active release from the formulation, combined by various methods that determine the concentration of active molecules in the receptor medium after a given time period [77,139]. The mechanism of release could be studied by fitting the in vitro release data to various mathematical models, of which Korsmeyer-Peppas release model (1) is often used. It describes the relationship between the fraction of active ingredient released (Mt/M∞) and the release time (t), using the release rate constant (k). The analysis takes into account the order of kinetics (n) and the constant k, assessing how close the release data are to the theoretical model [35,48]. Other methods that are used include spectral imaging, e.g., confocal microscopy, which produces high-resolution images that visualise the site of skin penetration after the administration of MNs, including the geometry of MNs [84,131].Confocal Raman spectroscopy (CRS) is an optical method that combines the principles of confocal microscopy and Raman spectroscopy for the study of in vivo penetration of topically applied substances. In a recent study [147], the skin penetration of niacinamide from various formulations was tested in vitro by the conventional Franz cells and in vivo using a quantitative CRS method under finite dose conditions; high correlation was found between the results. Krombholz and Lunter [148] have also applied the CRS method to the skin penetration of caffeine, showing the potential of achieving in situ measurement via CRS.Another novel in vitro analytical approach, Time of Flight—Secondary Ion Mass Spectrometry (ToF-SIMS) imaging, was recently used in the testing of cosmetics, and shown to be capable of measuring multiple compounds in the skin, simultaneously and in one single experiment [149]. This would be a very useful technique for the in vivo assessment of the efficacy of 3DP platforms.3D printed patches have simple geometry in general, such as film and grid/mesh. The earliest attempt to 3D-print a patch was the work by Goyanes et al. [27], with salicylic acid, an effective lipophilic keratolytic agent for acne treatment. A personalised nose-shape patch was created to provide complete occlusion and facilitate skin delivery of salicylic acid. The in vitro release of active from a standard formulation and a 3DP patch was performed using Franz diffusion cell. The initial release was slow, as expected for all delivery platforms. However, the SLA printed patch has shown a faster diffusion rate and a higher amount of total drug released within 3 h. An enhancement in salicylic acid delivery was observed by all 3DP types of patches in comparison to the standard formulation [27].The follow-up study by the same group was performed to investigate the personalised wound dressing made by 3D scanning followed by FDM and containing a series of antimicrobial agents. In line to the intended purpose, the release of actives from wound dressings has been reported as a dissolution rate (not the amount), since the skin barrier of the wound was disrupted [28]. It was observed that a higher active loading has resulted in a higher release from the 3D printed delivery platforms, since there was less hindering by the polymer matrix. However, this conclusion is not applicable for actives that are chemically bonded to the ink material. Therefore, studying chemical interactions between the 3DP platform matrix and the loaded actives is necessary for better understanding of the release mechanisms and further optimisation.3D printed wound dressings are the most studied patch-type skin delivery platforms. Biopolymers and hydrogels are the materials that have been tested for their printability by 3DP technologies. It was shown that rheological properties and swellability are the key features that determine the release of actives [105,116,137,150]. 3D printing of emulsions into grids was performed by Sommer et al. [90], whereby a hydrophobic dye, unisol blue, was incorporated as model active. It is worth noting that the emulsion has the potential to incorporate and simultaneously deliver hydrophilic and hydrophobic actives to the skin [90].Antioxidants are essential in anti-aging products, but also present in topical pharmaceuticals, hence they are the actives of interest for the 3DP platforms. One of them is lignin, an abundant natural antioxidant and antimicrobial ingredient, which was used in the 3DP patch intended for wound healing [117]. The printable filament was obtained by coating the PLA pellet with lignin solution, before applying hot melt extrusion. The resulting 3D printed patch was obtained by FDM. The product showed lower resistance to fracture and higher wettability than the one without lignin. In vitro release results showed that the delivery of the model drug, curcumin, was prolonged when loaded onto the 3D printed lignin-PLA mesh [117].Recently, Wang et al. [29] developed a 3D printed hydrogel patch with cryptotanshinone (CPT), a quinoid diterpene with anti-acne effect. To counteract its poor water solubility, CPT was entrapped in non-ionic surfactant–based vesicles, niosomes, before dispersing them into the hydrogel for 3D printing [29]. A comparison of in vivo permeation results was performed between an active-free 3DP patch, a conventional CPT-loaded patch, and a 3DP CPT-loaded patch, in addition to a commercial adapalene product, as a positive control. The results showed that the 3DP patch has shown better hydration and anti-acne effect, with CPT being delivered to deeper layers of the skin, indicating a positive effect of niosomes. However, there was a lack of direct comparison of the release results between the gel formulation with CPT-loaded noisomes and the 3DP patch with niosomes, therefore, it is hard to validate that the 3DP printed patch itself was the reason for the improved results [29].

There is enough evidence to conclude that 3DP technology has the potential to increase the efficacy of topically applied actives, specifically in the form of microneedles, where two mechanisms combine to enhance skin delivery: occlusion and the formation of micro-punctures. Hence, it would be worth exploring the possibilities within the cosmetic field more extensively.

8. Conclusions and Future Prospects

This review presents an overview and a critical assessment of the principles and main types of 3DP technology, production and characteristics of two key 3DP skin delivery platforms (patches and microneedles), and topical active ingredients used in these platforms. It considers relevant developments in a range of borderline areas, specifically in the medical area, aiming to inform and inspire prospective applications of 3D printing in cosmetics.

Based on the exponential growth in research on 3D printed MN technologies and the numerous associated advantages of this approach, it is envisaged that 3D printed MNs would gain prevalence over patches for both pharmaceutical and cosmetic applications. Dissolving MNs have shown the highest potential of the various MN types investigated so far, especially taking into account that comfort and convenience present major factors for consumer acceptance of cosmetic products.

Giving the limitations of conventional skin delivery systems [46,154], 3D printing might be able to provide personalised solutions. On the other hand, there are still difficulties in developing 3D printed skin delivery platforms, including the delivery of large molecules, skin irritation related to both MNs and patches, incomplete control of drug/cosmetic active release, and sustainability issues related to the raw materials used in 3DP. Due to the high cost of 3DP, when developing 3D printed skin delivery platforms, the priority will probably be given to personalised systems over mass production.

Although intense research has been carried out in the field of medical and pharmaceutical use of 3D printed microneedles during the last decade, the same does not apply to the field of cosmetic delivery. This indicates a large potential in developing efficient cosmetic delivery platforms using 3DP technology.