Self‐Seeded Nucleation of PET in a Benign Solvent Yields a High Modulus Aerogel With Ultra‐Low Thermal Conductivity Authors: Kira R. Baugh, Garrett F. Godshall, Glenn A. Spiering, Isabela Trindade Coutinho, Alejandro A. Rodriguez, Katherine J. Wood, Aiyana Gonzalez‐Arellano, Sadeq Malakooti, Stephanie L. Vivod, Robert B. Moore Published: 2026 ABSTRACT ABSTRACTPoly(ethylene terephthalate) (PET) is the most recycled plastic; however, upcycling of this commodity polymer to value‐added products has remained limited. Recent attempts to produce PET aerogels have relied on toxic solvents, chemical conversions and cross‐linking, and reinforcement to achieve useful mechanical performance. Now, using a new benign solvent (1,3‐diphenylacetone), we report a simple, safe, and sustainable dissolution and gelation procedure to convert waste PET into low density, monolithic aerogels with high mechanical strength (E = 20 ± 2 MPa) and remarkably low thermal conductivity (k = 21.9 to 28.9 mW m−1 K−1). The key to this process is controlled crystal nucleation by a self‐seeding phenomenon during gelation that establishes a mesoscale fibrillar framework. Our strategy offers a benign sol–gel process to upcycle plastic waste into a new form of pure PET for advanced thermal insulation. Introduction Aerogels are commonly recognized as low density, mesoporous, solvent extracted gels composed of a nano‐scale framework that exhibit solid‐like properties, high porosity, high internal surface areas, and often very low thermal conductivities [1]. Emerging applications of aerogels include lightweight materials for thermal and sound insulation [2, 3], catalyst supports [4], air filters [5], chemical absorbents in environmental remediation [6], and supports for controlled drug delivery [7]. Aerogels are typically formed through a sol–gel process yielding a substantially dilute, three‐dimensional interconnected framework that entraps solvent molecules (i.e., a gel), followed by a solvent extraction process that replaces the solvent with air, while retaining the original volume of the solvent‐swollen gel. Since the early introduction of fragile, inorganic, silica‐based aerogels by Kistler in 1931 [8], the field has grown significantly in recent years to include a wide variety of aerogel systems based on ductile organic polymers [1, 2]. Given the wide diversity of chemical functionality and segmental associations available with organic polymers, the macromolecular frameworks within polymeric aerogels have been designed and developed from both chemically cross‐linked covalent networks as well as thermo‐reversible physical networks formed through non‐covalent aggregation of chain segments. As a state‐of‐the‐art example, NASA has recently developed a number of lightweight, high modulus, covalently crosslinked polymer aerogels with high surface areas and low thermal conductivities, comparable to silica aerogels, for advanced aerospace insulation [9, 10]. Given the growing interest in polymer aerogels and the critical demand within our society for sustainable materials and processes, recent attention has focused on producing polymer aerogels from biomaterials [11, 12], benign solvents [13, 14], and plastic waste [1, 15, 16]. This attention has generated a timely interest in upcycling of waste poly(ethylene terephthalate) (PET) in the form of value‐added materials [17]. Recently, a chemical recycling process was used to create PET‐based aerogels, involving a chemical depolymerization of PET prior to gelation via chemical crosslinking [16]. Other efforts developed microporous, low density materials produced from mats of recycled PET fibers or PET particles that were chemically‐modified and/or cross‐linked with glutaraldehyde and poly(vinyl alcohol), exhibiting thermal conductivities comparable to expanded polystyrene (35 – 38 mW m−1 K−1) [18, 19, 20, 21, 22, 23, 24, 25, 26, 27]. These materials, while effective at repurposing waste PET, all contain components other than pure PET, and the chemical modifications (e.g., crosslinking) required to fabricate/stabilize the porous structures preclude further recycling of the PET component. Clearly, a true PET aerogel, fabricated using a benign sol–gel process, without any other components or chemical modifications to the polymer feedstock would offer significant advantages to the circular economy of waste PET. Gelation from solutions of semicrystalline polymers is often obtained using a temperature‐induced phase separation (TIPS) process [28, 29]. To date, dissolution and gelation of PET by TIPS has required the use of toxic and/or banned solvents (e.g., trifluoroacetic acid/dichloromethane mixtures). PET gels obtained from these solvents were typically found to be composed of poorly interconnected spherulites (i.e., radially ordered aggregates of crystalline lamella) exhibiting poor mechanical stability after solvent extraction [30, 31, 32, 33, 34]. As an advancement over these previous attempts, we have recently discovered that PET (e.g., sourced from waste water bottles) can be readily dissolved and gelled using TIPS from a new high boiling point solvent (i.e., 1,3‐diphenylacetone, DPA) that is non‐hazardous and listed as a US FDA approved food grade additive [35]. By carefully controlling the dissolution conditions that yield a high crystal nucleation density during gelation (i.e., self‐seeding), we have created a new, monolithic PET gel morphology consisting of a strong fibrillar framework of interconnected lamellar aggregates within the solvent‐swollen gel. Advantageously, the high nucleation density also prevents the undesirable formation of spherulites. Thus, following solvent extraction with supercritical CO2, a lightweight, semicrystalline aerogel of pure PET is formed that can sustain high compressive loads without any chemical modifications or other reinforcing components. Moreover, this simple and benign upcycling process yields a mesoporous PET aerogel with a high specific surface area, and remarkably low thermal conductivity that is ideal for advanced thermal insulation. Results and Discussion Fabrication of Monolithic PET Aerogels Monolithic PET aerogels of the present study are fabricated from flakes of waste water bottles through a TIPS process in a compatible, benign solvent, 1,3‐diphenylacetone (DPA). The value of the PET/DPA Flory‐Huggins interaction parameter at 210°C is χ = 0.24 (calculation in Note S1), which is characteristic of a good solvent and expected to promote TIPS through a solid‐liquid (S‐L) phase separation mechanism, leading to the formation of a volume‐filled semicrystalline gel [13, 14, 29]. As further support for the S‐L mechanism, the onset cloud point and crystallization temperatures are comparable and follow a similar trend with increasing polymer volume fraction ϕ PET (Figure S1) [29]. Given the favorable polymer‐solvent interactions, DPA promotes dissolution of PET at or above 210°C to yield a clear, homogeneous solution (Figure 1A) without any detrimental loss of molecular weight (MW) during the entire process (Figure S2). Upon cooling of these solutions to a temperature below the solvent‐depressed T m of PET (Figure 1B), PET chain segments begin to crystallize throughout the solution. Over just a few minutes, the polymer‐solvent mixture solidifies, whereby the semi‐crystalline microstructure entraps the solvent in an organogel (Figure 1A). Solvent exchange with ethanol, followed by supercritical CO2 drying yields the final pure PET aerogel as a lightweight, rigid monolith with the same total volume as the previous solvent‐swollen gel (Figure 1A). As these aerogels contain only PET and the high MW of the bottle‐grade polymer is maintained, it is conceivable that these upcycled materials could be later recycled if desired. Specifically, the TIPS and solvent extraction processes used to make these aerogels offer purification by recrystallization, conceivably extracting low MW impurities and additives, to yield a purified bottle grade feedstock (Figure S3). Simple fabrication of PET aerogels is accomplished with a benign solvent. (A) Waste PET water bottle flakes are dissolved in DPA at elevated temperatures and cooled to form a solid gel. Following solvent extraction, a lightweight, mechanically robust, PET aerogel monolith is formed. Considering lifecycle circularity, the purified PET aerogel can be further repurposed, and the DPA extracted from the polymer gel can be reclaimed from ethanol and reused. (B) A schematic sol–gel diagram representing the ideal thermal conditions needed to form stable PET aerogels (i.e., T dis < T m, offset and T gel ≤ T c, onset). The onset and offset temperatures for melting and crystallization were measured by differential scanning calorimetry (DSC) (Figure S4) and extrapolated to the red and blue shaded regions in the diagram. (C) The melting and crystallization phase diagram of PET/DPA mixtures, data collected using DSC. Upon heating, the PET/DPA gel melts over the range T m, onset < T m, peak < T m, offset (Figure S4A). Upon cooling, the PET/DPA sol solidifies at T gel ≤ T c, onset (Figure S4B). The melting and crystallization phase diagram for PET/DPA solutions is shown in Figure 1C. Upon heating, the PET crystallites within the gel melt over a range of temperatures between T m, onset and T m, offset with a peak melting temperature, T m, peak, that depends on the polymer volume fraction, ϕ PET, according to Flory's equation [36]: (1) 1Tm,peak−1Tmo=R·Vm,pΔHfo·Vm,sϕs−χϕs2 where T m, peak is the experimental peak of the melting temperature curve of the PET/DPA gel, Tmo is the equilibrium melting temperature of PET (Tmo = 258°C) [37], R is the universal gas constant, V m,p is the molar volume of the polymer repeat unit (V m,p = 139.1 cm3 mol−1), ΔHfo is the enthalpy of fusion per mol of repeat units for an infinitely thick PET crystal (ΔHfo = 26.9 kJ mol−1) [38]. V m,s is the molar volume of the solvent (V m,s = 196.7 cm3 mol−1) [13], ϕ S is the solvent volume fraction, and χ is the Flory‐Huggins interaction parameter for PET in DPA (χ = 0.24). The broad range of melting temperatures between T m, onset and T m, offset are typical for semi‐crystalline polymers and attributed to melting of chain folded lamella having a wide range of lamellar thicknesses [39]. Upon cooling, the PET component in the solution crystallizes over a similarly wide range of temperatures between T c, onset and T c, offset. The specific dissolution temperature used for the 15 wt.% PET/DPA mixture (i.e., T dis = 210°C, Figure 1B) was intentionally chosen to be just below the T m, offset = 230°C. We have discovered that this incomplete melting condition (i.e., melt memory) yields a visually clear solution (Figure 1A); however, the remaining, persistent nuclei are critical in achieving a self‐seeded nucleation during the subsequent crystallization during gel formation. Effect of Dissolution Temperature on PET Aerogel Morphology We have found that the development of the critical morphological features constituting the structural framework of these PET aerogels is strongly dependent on the dissolution temperature used to prepare the original 15 wt.% PET/DPA solutions. When the PET flakes are dissolved in DPA at a temperature of 210°C (Figure 2A) and isothermally gelled at temperature T gel ≤ 140°C a homogeneous, monolithic structure is formed, constituting an isotropic network of interconnected nanofibrils and mesopores after solvent extraction (Figure 2B,C). Gels formed under these conditions are solid, firm, and stable to stresses encountered during solvent exchange and extraction. In stark contrast, a slightly higher dissolution temperature of 230°C (Figure 2D; i.e., equal to the T m, offset) yields a gel morphology commonly observed in the literature [30, 31, 32, 33, 34], comprising a heterogeneous collection of poorly connected large globular features of uniform size (∼ 10 µm in diameter), as shown in Figure 2E. These gels are soft and paste‐like, often disintegrating during solvent exchange. Dissolution temperature controls the morphology of PET aerogels. (A) Schematic diagram of the fibrillar (T dis = 210°C) TIPS aerogel processing conditions. (B,C) SEM micrographs of a 15 wt.% PET aerogel processed at a dissolution temperature of 210°C reveal a monolithic morphology composed of isotropically oriented and highly interconnected nanofibrils. (D) Schematic diagram of the spherulitic (T dis = 230°C) TIPS aerogel processing conditions. (E) SEM micrograph of a 15 wt.% PET aerogel processed at a dissolution temperature of 230°C reveals a globular morphology with large, disconnected structures. (F,G) Crossed‐polarized PLM and SALLS of PET/DPA gels cooled from a dissolution temperature of 210°C show no evidence of birefringence. (H,I) Crossed‐polarized PLM and SALLS of PET/DPA gels cooled from a dissolution temperature of 230°C confirm that the globular features are typical birefringent spherulites. (J) USAXS‐SAXS‐WAXS of 15 wt.% PET aerogels, comparing the fibrillar and spherulitic scattering features. Profiles are offset for visualization purposes. (K) DSC cooling thermograms of PET/DPA gels reveal crystallization exotherms that are dependent on dissolution temperatures ranging from 195°C to 230°C. (L) The plot of peak crystallization temperature versus T dis delineating Domain I and II behavior. The dashed lines indicate the three dissolution temperatures compared in this study. Thin sections of the PET/DPA solutions were examined by cross‐polarized light microscopy (PLM) and cross‐polarized small‐angle laser light scattering (Hv SALLS). For the PET/DPA solution prepared at 210°C, cooling to T gel ≤ 140°C produced a featureless (dark) PLM image and negligible Hv SALLS scattering, indicating no birefringence (Figure 2F,G) [40, 41]. In contrast, the PET/DPA solution cooled from 230°C, yielded a distinct four‐lobe pattern in Hv SALLS (Figure 2H), and classic Maltese cross extinction patterns in PLM (Figure 2I), characteristic of birefringent polymer spherulites with radially oriented lamella [42]. The average spherulite size as measured by PLM (10.9 ± 0.4 µm) agrees with the size of globular features observed by SEM (Figure 2E), now confirmed as spherulites. Given that the PET component in both the spherulitic and fibrillar gels is semi‐crystalline with comparable degrees of crystallinity (X c = 42.8 ± 3.5% and 43.7 ± 5.4%, respectively), the lack of birefringence in the fibrillar gel suggests that the crystalline lamella are aligned along the major axis of the fibrils. Since the nanofibrils are isotropically oriented, there are no preferential optical axes in the bulk fibrillar gel, and thus no birefringence. The hierarchical morphology of the solvent extracted PET aerogels was characterized using ultra small‐, small‐, and wide‐angle X‐Ray scattering (USAXS/SAXS/WAXS) (Figure 2J). The scattering patterns for both gel morphologies display several common features consistent with the hierarchical structure of semicrystalline polymer aerogels [13, 14, 43]. Sharp wide‐angle diffraction peaks beyond q = 10 nm−1 are attributed to reflections consistent with the triclinic crystal structure of PET [44], confirming an identical crystal structure in both gel systems. The broad SAXS knee between q = 0.3 nm−1 and 0.7 nm−1 is associated with a dimension characteristic of the PET crystalline lamellae [45, 46]. Using the unified fit approach of Beaucage [47], these aerogels were found to contain lamella with thicknesses, l c, on the order of 10 nm. Distinct oscillations in the scattering profiles are also observed for both systems beginning at q = 0.5 nm−1 and persisting to q = 4 nm−1, and are attributed to a periodic structure factor associated with locally ordered stacks of lamella within both fibrils (i.e., axialites) and spherulites [14, 45, 46]. The radius of gyration of the PET crystallite feature was found to be between 5.5 and 6.5 nm depending on morphology, and the Porod exponent for the crystallite feature was found to be 4.0, consistent with smooth interfaces of crystalline lamellae (Table S1). The scattering profiles for the fibrillar and spherulitic aerogels differ significantly in the USAXS region. For the spherulitic aerogel, an abrupt increase in intensity is observed at q values below 0.007 nm−1. This additional scattering feature is attributed to the large (micron‐scale) dimensions of the ordered spherulites. While the relatively large inter‐fibrillar pores observed in the SEM images (Figure 2C) account for the majority of the porous volume of these aerogels, nitrogen porosimetry primarily probes the nanometer‐size pores (e.g., mesopores) [48, 49]. The porous fibrillar aerogels were found to have a surface area of 167 m2 g−1 compared to a ∼2x lower surface area of 86.8 m2 g−1 for the denser spherulitic aerogels. Consistent with a less dense framework, the fibrillar aerogels revealed a larger population (by volume) of mesopores compared to the denser spherulitic aerogels. The nitrogen absorption isotherms (Figure S5A) revealed that both the fibrillar and spherulitic aerogels yield an IUPAC type IV isotherm, characteristic of a mesoporous absorbent with 2–50 nm pores (Figure S5B) [50]. Moreover, the sickle‐shaped hysteresis profile between 0.7 and 1.0 p/p is classified as an IUPAC type H3 hysteresis, characteristic of capillary condensation in slit‐like mesopores formed from aggregates of platelike particles. This classification is consistent with nanometer‐scale pores between stacked crystalline lamella within the fibrils and spherulites of the solvent extracted aerogels [13, 14]. Impact of Melt‐Memory on PET Aerogel Morphology and Mechanical Properties Following the self‐seeding approach of Fillon [51], DSC was used to probe the effect of dissolution temperature, T dis, on the crystallization/gelation behavior of PET/DPA solutions upon cooling, as shown in Figure 2K,L. The PET/DPA gels were first heated to temperatures at which the gels “dissolved” (i.e., melted or partially melted in solution) ranging from 195°C to 230°C and then cooled at −10°C min−1. As shown in Figure 2K, the samples exhibited crystallization exotherms that shifted to lower temperatures with increasing T dis. A plot of the peak crystallization temperature, T c, peak, versus T dis (Figure 2L) reveals that T c, peak decreases with increasing T dis over the range 195°C < T dis < 218°C, but remains constant at T dis > 218°C. This behavior is common for crystallizable polymers and is characteristic of a self‐nucleation phenomenon referred to as “melt‐memory” [28, 51, 52]. At relatively low dissolution temperatures, a small but significant fraction of the polymer chains in “solution” retain some degree of crystalline order, and these residual, persistent crystallites (self‐seeds) provide a high nucleation density capable of initiating early crystallization upon cooling (Domain II behavior). At higher dissolution temperatures, all crystalline order in the solution is eliminated leaving only a low heterogeneous nucleation density and thus a constant T c, peak upon cooling (Domain I behavior). As shown in the SEM micrographs of Figure 2, the PET aerogel morphology is profoundly impacted by the melt‐memory phenomenon. Monolithic, fibrillar aerogels are generated by selecting T dis = 210°C, just below the point at which the last PET crystals melt (i.e., T m, offset = 230°C; Figure 1C). With incomplete melting (Domain II, low T dis), the self‐seeds establish a high nucleation density that crowds the growth of the crystalline domains. This crowding effect limits the crystalline development to small axialites, resulting in a fibrillar gel morphology. In contrast, a spherulitic morphology is generated by selecting T dis = 230°C, where all self‐seeds are eliminated. With a low nucleation density (Domain I, high T dis), the crystalline domains have room to grow and organize into well‐developed spherulites. Thus, by carefully controlling the dissolution temperature, and thus the nucleation density, the aerogel morphology can be tuned to yield a robust physical network of semicrystalline fibrils presumably interconnected by amorphous tie chains. Moreover, this control of nucleation density provides an effective process to eliminate the commonly observed formation of discrete spherulites that result in weak gels with poor inter‐spherulitic connectivity across their boundaries. The mechanical performance (in compression) of the monolithic fibrillar aerogels is compared to the spherulitic aerogels in Figure 3A. At low strains, the fibrillar aerogel monolith (Figure 3B, density ρ b = 0.173 g cm−3) exhibits a yield stress of 0.515 MPa and a compressive modulus of 20 ± 2 MPa. This modulus value is significantly higher than conventional silica aerogels and comparable to commercial, chemically crosslinked polyurea aerogels (Airloy X103) of a similar density (Figure 4E) [53]. Beyond the yield point, the compression profile transitions to a plateau of plastic deformation (at near constant stress) attributed to irreversible buckling and reorientation of the fibrillar framework (Figure 3G), similar to behavior observed for an open cell foam [54, 55]. This plateau extends to about 60% strain until the onset of densification, where the pores collapse and the fibrils compact upon themselves leading to a compressed flat platelet at the end of the test (Figure 3C), In stark contrast, the spherulitic aerogel samples processed at T dis = 220°C experienced brittle failure (Figure 3D) with a load of only 0.06 MPa. This poor compressive performance is attributed to weak connectivity between the disordered spherulites (Figure 3H), which facilitates crack propagation. The PET gels processed at T dis = 230°C were not tested, as they disintegrated into powders upon solvent extraction (Figure 3E) due to the presence of highly ordered spherulites (Figure 3I). Based on the relatively high yield stress of the fibrillar PET aerogel, a 10.2 cm × 10.2 cm × 0.64 cm slab can support approximately 1200 lbs. force and thus proves to be remarkably robust and resistant to compressive failure while supporting the weight of mid‐size sedan (Figure 3F and Movies S1 and S2). The mechanical robustness of PET aerogels is dependent on their morphology. (A) A comparison of the compressive stress‐strain curves of the mechanically robust, fibrillar PET aerogel (T dis = 210°C) and the brittle, spherulitic PET aerogel (T dis = 220°C). (B) Uncompressed 15 wt.% PET aerogel cylinder prior to compression testing. (C) The PET aerogel processed at T dis = 210°C compresses into a flat platelet at the completion of the compression test. (D) The PET aerogel processed at T dis = 220°C suffers from catastrophic brittle failure during the compression test. (E) The PET aerogel processed at T dis = 230°C is incapable of forming a cylindrical test specimen and disintegrates into a powder upon solvent extraction. (F) Sedan rollover test of a 10.2 cm × 10.2 cm × 0.64 cm 15 wt.% PET aerogel sample (see Movies S1 and S2 for the test). (G through I) SEM micrographs of PET aerogels processed with dissolution temperatures of 210°C, 220°C, and 230°C, respectively. Monolithic PET aerogels exhibit remarkably low thermal conductivity. (A) An illustration of the Knudsen effect in fibrillar PET aerogel monoliths, where the pore size is sufficiently small as to inhibit the transfer of gas molecules with a mean free path of ∼65 nm through the material. (B,C) A large format 10.2 cm × 10.2 cm × 0.64 cm 15 wt.% PET aerogel monolith viewed in the visible spectrum. (D) Thermal conductivity measurements of the PET aerogel using the calibrated hot plate method and a ThermTest HFM‐25 heat flow meter. Thermal conductivity is compared to other thermal insulating materials of comparable density and compressive modulus. The thermal conductivity of the conventional silica aerogel is as reported in reference #53 (Guarded Hot Plate method) [53]. (E) The FLIR image highlights the contrast between the hand at body temperature and the room temperature background. (F) FLIR image of the same 15 wt.% PET aerogel in 5A, shown to effectively mask the thermal signature of a human hand (i.e., IR stealth). Thermal Conductivity of PET Aerogels Aerogels are often ideal thermal insulators, due to their ultra‐low thermal conductivity [1, 11]. These remarkable thermal properties are attributed (in part) to the Knudsen effect, where gas phase heat transfer is greatly reduced when the gas is confined within nanometer scale pores (Figure 4A) [56]. Since the mesopores in the PET aerogel are in the range of 2 to 60 nm (Figure S5B) and are smaller or comparable to the mean free path of gas molecules (∼ 65 nm) [56, 57], infrequent collisions between gas molecules within a given pore limit the gaseous heat transfer, thus lowering the gaseous thermal conductivity. Thermal conductivity through the solid phase of the aerogel is also affected by porosity and tortuosity. The 15 wt.% fibrillar PET aerogels were found to have a bulk density of ρb = 0.173 g cm−3 and a skeletal density of ρs = 1.40 g cm−3, leading to a porosity of Π = 87.6% (calculation in Note S2). The highly porous nanostructure of the fibrillar PET aerogels (Figure 2C), coupled with a high surface area (167 m2 g−1, BET method) yields a highly tortuous solid structure that inhibits the solid heat transfer pathway (Figure S5A). Figure 4B through Figure 4F show a large format 15 wt.% PET aerogel slab held in the palm of a hand. When the hand alone is viewed through a forward‐looking FLIR camera (Figure 4C), there is significant thermal contrast between the hand (Figure 4E, 37°C) and the room temperature environment. When the PET aerogel slab is placed on the hand and viewed through the FLIR camera, the aerogel blocks the transmission of thermal radiation and appears thermally opaque (Figure 4F). The ability of the mesoporous PET aerogel to effectively mask the thermal signature of a human hand is attributed to its low thermal conductivity in a phenomenon often referred to as infrared stealth [58]. Thermal conductivity measurements of highly insulating aerogel materials is recognized to be challenging and often dependent on the specific method used and experimental conditions [59]. To facilitate comparisons to other measurements in the literature, two steady‐state methods (calibrated hot plate and heat flow meter) were used to measure the thermal conductivity of the 15 wt.% PET aerogel (shown in Figure S6 and described in Section 4). With the calibrated hot plate and heat flow meter methods, the PET aerogel was found to have an ambient thermal conductivity of 28.9 ± 0.9 and 21.9 ± 0.04 mW m−1 K−1, respectively (Figure 4D). When tested at −50°C using a modified transient plane source method (Trident, C‐Therm Technologies), the thermal conductivity of the PET aerogel was found to be 20.8 ± 0.2 mW m−1 K−1. This remarkably low thermal conductivity is comparable to the previously reported thermal conductivity of a commercial polyurea aerogel (Airloy X103) and a state‐of‐the‐art polyimide aerogel from NASA (Figure 4D) [10], and is significantly lower than that of similar density expanded polystyrene (NIST SRM 1453) [60]. While the thermal conductivity of the polymeric aerogels is higher than that of a conventional silica aerogel [53], the compressive modulus of polymer aerogels is roughly an order of magnitude higher (Figure 4D). To the best of our knowledge, this is the lowest thermal conductivity reported for any PET‐based aerogel. Conclusion Previous attempts to create PET aerogels have required the use of toxic solvents, other polymeric binders, and/or chemical modifications and crosslinking to achieve suitable physical properties. With our discovery of a new benign solvent for PET, we have designed a new sustainable approach to produce pure PET aerogels (without any other components) from waste water bottles that exhibit superior mechanical and thermal properties compared to previous PET aerogel attempts. With a simple TIPS process, we have discovered that the morphology of the resulting aerogel can be tuned by precise control of PET crystal nucleation during the gelation process. By selecting a dissolution temperature in DPA that is just below the highest solvent‐depressed melting point of the PET crystallites, the incomplete melting preserves a significant quantity of residual crystallites in the polymer/solvent mixture that act as efficient seeds for prolific nucleation needed to develop a monolithic fibrillar physical framework capable of supporting significant mechanical load. The mesoporous structure of the nanoscale fibrillar PET aerogels yields remarkably low thermal conductivity comparable to state‐of‐the‐art, chemically crosslinked polyimide aerogels. The ultra‐low thermal conductivity, combined with our benign process for repurposing single use waste PET packaging, provides a value‐added, long lifetime application as an aerogel for advanced, lightweight thermal insulation. With respect to circularity in the life cycle of plastics, this benign gelation and solvent extraction process is fundamentally a purification by recrystallization process that preserves the high molecular weight of the bottle grade PET for subsequent repurposing applications if desired. Experimental Section/Methods Materials PET flakes were made by cutting waste 500 mL water bottles (Deer Park). 1,3‐diphenylacetone was purchased from Sigma–Aldrich (St. Louis, MO) and Oakwood Chemical (Estill, SC). Hexafluoroisopropanol was purchased from Sigma–Aldrich (St. Louis, MO). Ethanol (100% USP, 200 proof) was purchased from Sigma–Aldrich (St. Louis, MO). Argon (99.999% UHP) and carbon dioxide (99.5% pure, equipped with an eductor tube) were purchased from Linde Gas & Equipment (Richmond, VA). All materials were used as received. PET Aerogel Fabrication PET flakes (approximately 1 cm2 in area and 0.1 mm thick) and DPA were combined in glass test tubes (22 mm and 15.4 mm, ID) to constitute mixtures of 15 wt.% PET. The test tubes were then capped with a rubber septum and equipped with a glass stirring rod. These sealed tubes were purged with argon for a minimum of 15 min prior to heating, and the argon purge was continued throughout heating and dissolution. The test tubes were then placed in a pre‐heated Hart Scientific 9123/9122 well heater for large tubes and small tubes, respectively. The preset dissolution temperatures for the PET/DPA mixtures ranged from 210°C to 230°C with an accuracy of ± 0.1°C. Solutions were periodically stirred by hand with the glass stirring rod to promote PET dissolution in the solvent. Once the PET/solvent mixture was visibly homogeneous (observed after 15 to 30 min of heating), with no remaining particles suspended in solution, the test tube was removed from the well heater and placed into an oil bath equilibrated at a set gelation temperature (Tgel = 140°C) below T m, onset (Figure 1 and Figure S4), with an accuracy of ± 5°C. Over a period of time (seconds to minutes), depending on volume (between 2 to 5 min for 15 wt.% gels in a 15.4 mm ID glass test tube), the clear solution transformed into an opaque gel following the temperature induced phase separation (TIPS) process. After gelation, the gel was held at the crystallization temperature for 20 additional minutes to ensure crystallization throughout the full volume. The solidification by gelation phenomenon in TIPS is commonly classified by the mechanism of phase separation as either a liquid‐liquid (L‐L) [61] or a solid‐liquid (S‐L) [29] phase separation. These mechanisms are generally distinguished by the magnitude of Flory's polymer‐solvent Chi interaction parameter, χ [62, 63]. Upon cooling in a relatively poor solvent (i.e., χ > 0.5), (L‐L) phase separation occurs, followed by crystallization in the polymer rich phase. With more favorable polymer‐solvent interactions (i.e., χ < 0.5), the polymer simply crystallizes in solution upon cooling by a (S‐L) process once the temperature drops below the solvent‐depressed melting point (i.e., crystallization occurs without liquid‐liquid phase separation). Once fully solidified, the solvent swollen gel was extracted from the test tube and placed in a good solvent for DPA (100 mL of ethanol) for solvent exchange. The ethanol was replaced with fresh solvent every 24 h for 5 days to ensure complete removal of DPA from the gel. Gas chromatography (GC) confirmed that the solvent exchanged gel contained less than 100 ppm of DPA after the 5‐day protocol (Figure S3). Next, the solvent‐exchanged PET gel was dried with supercritical CO2, using a Leica EM CPD300 Critical Point Drying System. The larger PET aerogel samples used for thermal conductivity measurements and the car rollover test were dried using supercritical CO2 at NASA Glenn Research Center, per reference #10 [10]. Supercritical Drying of PET Gels A Leica EM CPD300 Critical Point Drying System was used to dry the small cylindrical samples. The method used to dry the gels is as follows: cool the chamber to 15°C, pressurize with CO2 at a slow speed setting, perform 7 full exchanges of the liquid CO2 in the drying chamber at speed setting 5 out of 10, heat to 41°C at a slow speed setting, and then release the CO2 in a supercritical state at the lowest speed setting possible. During the procedure, the stir bar is spinning at 50% of maximum speed, and CO2 exchange begins after a delay of 2 min following the first influx of CO2 into the drying chamber. Larger samples were dried in a custom‐built AccuDyne critical point dryer with large drying chambers (≥4L) [10]. After supercritical drying, the larger samples were vacuum dried at 50°C (below the Tg of PET) for approximately 2 h after being removed from the supercritical CO2 chamber to ensure the removal of any remaining ethanol. Differential Scanning Calorimetry DSC experiments were performed on a TA Instruments Discovery DSC2500. For the DSC tests at different ϕPET (Figure 1), solvated gels were ground into a powder and packed into hermetically sealed DSC pans. To determine crystallization temperature, samples were heated at 10°C min−1 to 300°C to erase thermal history, held isothermally for 3 min, cooled at 10°C min−1 to 40°C, and then heated 10°C min−1 to 300°C (Figure 1C and Figure S1). For the self‐nucleation tests (Figure 2), approximately 5 mg of solvated PET/DPA gel was packed into hermetically sealed DSC pans. Samples were heated at 10°C min−1 to 230°C, held isothermally for 3 min, then cooled at 10°C min−1 to 40°C to erase pre‐existing thermal history of the gel. These two steps were repeated to ensure each sample can be compared. Next, samples were heated to the desired dissolution temperature, heating at 10°C min−1 before immediately cooling back to 40°C at 10°C min−1. Data shown in Figure 2K,L. Optical Microscopy Optical microscopy was used to determine PET/DPA gel cloud point, and to determine the existence of spherulites in poor gelation conditions. For cloud point testing, solvated gels were ground into a fine powder and packed into 1.5 mm diameter Pyrex capillaries, which were subsequently flame‐sealed with a torch. These experiments were performed on a Nikon Eclipse LV100 optical microscope equipped with a Linkam HFSX350‐CAP hot stage outfitted with a 1.7 mm diameter capillary port. Conditions were monitored using an AmScope (MU503B) digital CMOS camera and AmScope software. To determine the cloud point, capillaries were heated in the hot stage to 300°C at 50°C min−1, held isothermally for 3 min, and cooled at 10°C min−1 to 40°C. Snapshots of the capillary were taken (one per second), and the brightness of each image was extracted using the ImageJ image analysis software. A plot of brightness versus temperature was constructed, and the cloud point temperature was taken as the onset of brightness decrease using TRIOS data analysis software. Three replicate samples were tested for each polymer/solvent composition (Figure S1). Polarized light microscopy (PLM) was used to identify the presence of spherulites in 15 wt.% PET/DPA gel samples made under different dissolution and gelation conditions. Two samples were prepared by placing a small amount of a 15 wt.% PET/DPA gel, previously dissolved at 210°C and gelled at 140°C, between two glass covers. The first sample was heated to 230°C and then cooled to 140°C, while the second sample was heated to 210°C and then cooled to 140°C, using a Linkam HFSX350‐CAP temperature stage to mimic different dissolution temperatures. The samples were imaged during this process, prior to being used for small‐angle laser light scattering measurements. These experiments were performed on a Nikon Eclipse LV100 optical microscope equipped with a Linkam HFSX350‐CAP hot stage and were monitored using an AmScope (MU503B) digital CMOS camera and AmScope software. Viscosity PET and PET aerogel particles were dissolved in hexafluoroisopropanol at various concentrations (0.2, 0.4, 0.6, 0.8, and 1 g dl−1). The PET was allowed to dissolve in the solution overnight prior to testing. Samples were then tested using a Cannon‐Ubbelohde Viscometer (9721‐K582) submerged in a 35°C water bath. Samples were allowed to equilibrate for 30 min to 35°C in the viscometer prior to testing. Results in Figure S2. Gas Chromatography Gas chromatography measurements were taken on an Agilent 7890A GC. A calibration curve of DPA in acetone was first prepared. Then, PET aerogel samples that had been solvent exchanged for the 5‐day period were placed back into fresh acetone to extract any remaining DPA. The resulting acetone was then tested for trace DPA (Figure S3). Scanning Electron Microscopy SEM samples were prepared by fracturing the aerogel at room temperature and imaging the fracture surface. Aerogel samples were sputter‐coated with a 5 nm‐thick layer of iridium using a Leica EM ACE600 sputter coater before imaging. Images were taken on a LEO (Zeiss) 1550 field emission scanning electron microscopy (FE‐SEM) with in‐lens detection. The image analysis software ImageJ was used to collect dimensional data on the microstructure of PET aerogels. Small‐Angle Laser Light Scattering Small‐Angle Laser Light Scattering (SALLS) were used to identify the presence of spherulites and measure their diameter. Two samples were prepared by placing a small amount of a 15 wt.% PET/DPA gel, previously dissolved at 210°C and gelled at 140°C, between two glass covers. The first sample was heated to 230°C and then cooled to 140°C, while the second sample was heated to 210°C and then cooled to 140°C, using a Linkam HFSX350‐CAP temperature stage to mimic different dissolution temperatures. The samples were then analyzed using SALLS. The scattering patterns were obtained in transmission mode with perpendicular polarizers (VH mode) using a laser font of a 3 mW He‐Ne laser (λ = 632.8 nm) as the light source. The sample‐to‐detector distance was 24 cm, and the scattering vector (q) range was calibrated using a 300 grooves per mm diffraction grating. The four‐leaf clover patterns characteristic of spherulites were acquired using an AmScope digital camera (MU503B). The software SAXSGUI was used to obtain radially integrated scattering intensity as a function of the scattering vector (q). The diameter (D) of the spherulites was obtained using the following equation [41]: (2) D=2·1.025·λn·π·sinθ2 where λ is the wavelength of the light source (632.8 nm), n is the refractive index of the sample and θ is the angle at which the scattering intensity is maximum. The refractive index of the sample was calculated according to the relationship in the equation below: (3) n=nPET·vol%ofPET+nDPA·vol%ofDPA where v% of PET and v% of DPA are the volume fraction of PET and DPA, respectively, and n PET is the refractive index of PET (1.56) and n DPA is the refractive index of DPA (1.57). X‐Ray Scattering X‐ray scattering samples were prepared by placing a 1 mm thick section of PET aerogel in between two pieces of Scotch Magic Tape, before mounting the sample to the solid sample plate in the SAXS. USAXS, SAXS, and WAXS experiments were performed using a Xeuss 3.0 USAXS/SAXS/WAXS beamline, equipped with a GeniX 3D Cu HFVLF microfocus X‐ray source with a wavelength of 0.154 nm (Cu Kα). The sample‐to‐detector distance was 1800 mm for extra‐small angle X‐ray scattering (ESAXS), 900 mm for SAXS, 370 mm for mid angle X‐ray scattering (MAXS), and 43 mm for WAXS. A Bonse‐Hart camera was used to collect USAXS data. The q‐range was calibrated using a lanthanum hexaboride standard for WAXS and a silver behenate standard for ESAXS, SAXS, and MAXS. Two‐dimensional scattering patterns were obtained using a Dectris EIGER 4 M detector, with an exposure time of 4 h for USAXS, 4 h for ESAXS, 2 h for SAXS, 1 h for MAXS, and 1 h for WAXS. The scattering data was reduced and corrected for background, thickness, and transmission using the XSACT software. Slit‐smeared USAXS data was desmeared using XSACT, yielding pinhole equivalent data. ESAXS, SAXS, MAXS, and WAXS data were output on an absolute scale using direct beam intensity. USAXS, ESAXS, SAXS, MAXS, and WAXS profiles were merged using the XSACT software. The scattering profiles were vertically shifted to facilitate comparison. Scattering plots are presented as scattering intensity, I(q), versus scattering vector q, (4) q=4πλ·sinθ where θ is one half of the scattering angle and λ is the X‐ray wavelength. Gas Adsorption Nitrogen porosimetry measurements were taken using a Micrometrics 3Flex gas sorption analyzer, testing via a physisorption method. PET aerogel samples were degassed at 65°C under nitrogen for at least 20 h before testing. Adsorption and desorption isotherms were collected using nitrogen as the adsorbent at −196°C (77 K). Surface area was calculated using the Brunauer–Emmett–Teller (BET) method [48], and pore size distribution was calculated using the Barrett–Joyner–Halenda (BJH) method [49]. Plots are on Figure S5. Compression Testing Compression testing was used to determine the impact of morphology on the strength of PET aerogels. Cylindrical aerogel samples were prepared in accordance with the 2:1 length‐to‐diameter ratio specified by ASTM D695‐23 (approximately 30 mm x 15 mm) [64]. Stress‐strain measurements (in compression mode) were obtained using an Instron Dual Column Table Model 68TM‐30 equipped with 2501 series compression platens and a 30 kN load cell. Samples were loaded at a rate of 1.3 mm min−1 until failure or after exceeding 90% compression. Yield strength was determined to be at the stress at which the stress–strain curve intersected a line with a slope identical to the linear region offset by 0.002 mm mm−1 strain. Yield strain was chosen as the x‐coordinate of the yield strength. Thermal Conductivity, Hot Plate Method Steady‐state thermal conductivity measurements were performed using a Calibrated Hot Plate Thermal Conductivity Analyzer developed by Aerogel Technologies Inc. (Boston, MA), following ASTM Standard E1225‐09 (Figure S6A) [65]. In this setup, a 100 mm × 100 mm × 0.25 mm, 15 wt.% PET aerogel sample (of unknown thermal conductivity) and a National Institute of Standards and Technology (NIST) reference material with known thermal conductivity (expanded polystyrene board, NIST SRM 1453) [60] are placed in series between a hot plate (set at 37.5°C) and a cold plate (set at 0°C). Assuming one‐dimensional heat transfer under steady‐state conditions, the thermal conductivity of the aerogel (k aerogel) is calculated using the following equation [1]: (5) kaerogel=kNIST·ΔTNISTΔTaerogel·taerogeltNIST where, k NIST is the thermal conductivity of the reference material (NIST SRM 1453, with a bulk density of 0.048 g cm−3 and a room temperature thermal conductivity of 35.6 mW m−1 K−1), and ΔT NIST, ΔT aerogel, t NIST, and t aerogel are the temperature differences and the thicknesses of the NIST standard sample and the aerogel specimen, respectively [60]. Thermal Conductivity, Heat Flow Meter Method Steady‐state thermal conductivity measurements were also performed using a ThermTest HFM‐25 Heat Flow Meter, on a 100 mm × 100 mm × 0.25 mm, 15 wt.% PET aerogel sample (Hanwell, NB, Canada) (Figure S6B). The sample was tested according to ASTM Standard C518‐21 [66]. The HFM‐25 is calibrated for aerogels with NIST SRM 1453 [60]. Assuming one‐dimensional heat transfer under steady‐state conditions, the thermal conductivity of the aerogel (k aerogel) is calculated using the following equation: (6) kaerogel=S·E1+S·E22·taerogelT1−T2 where S is the calibration factor of the heat flux transducers, E1 and E2 are the heat flux outputs for two adjacent transducers, t aerogel is the thickness of the aerogel specimen, and T1 and T2 are the hot and cold plate temperatures, respectively. Thermal Imaging Infrared images were collected on a Teledyne FLIR E53‐24 Advanced Thermal Imaging Camera. PET aerogels were prepared in accordance with the procedure outlined in this work. Funding Preliminary investigations leading to the discovery of PET gelation in DPA were supported the National Science Foundation under the award 1809291. Conflicts of Interest The authors (KRB, GFG, RBM) declare an international patent application (PCT) #WO2026055420A1 has been filed covering the PET aerogel formation described in this manuscript. Supporting information Supporting File: adma73409‐sup‐0001‐SuppMat.docx. Supporting File: adma73409‐sup‐0002‐MovieS1.mp4. Supporting File: adma73409‐sup‐0003‐MovieS2.mp4.