Carbon 208 (2023) 148–159 Contents lists available at ScienceDirect Carbon journal homepage: www.elsevier.com/locate/carbon Lithium halide filled carbon nanocapsules: Paving the way towards lithium neutron capture therapy (LiNCT) Gil Gonçalves a, b, **, Stefania Sandoval a, Marina Llenas a, Belén Ballesteros c, Tatiana Da Ros d, Silva Bortolussi e, f, Laura Cansolino f, g, Cinzia Ferrari f, g, Ian Postuma f, Nicoletta Protti e, f, Manuel Melle-Franco h, Saverio Altieri e, f, Gerard Tobías-Rossell a, * a Institut de Ciència de Materials de Barcelona (ICMAB-CSIC) Campus de la UAB, 08193, Bellaterra, Spain b TEMA, Mechanical Engineering Department, University of Aveiro, 3810-193, Aveiro, Portugal c Catalan Institute of Nanoscience and Nanotechnology (ICN2), CSIC and the Barcelona Institute of Science and Technology, Campus UAB, Bellaterra, 08193, Barcelona, Spain d INSTM - Trieste unit - Department of Chemical and Pharmaceutical Sciences, University of Trieste, Via L. Giorgieri 1, 34127, Trieste, Italy e University of Pavia, Department of Physics, Pavia, Italy f Istituto Nazionale di Fisica Nucleare (INFN), The Unit of Pavia, Pavia, Italy g University of Pavia, Department of Clinic-Surgical Sciences, Experimental Surgery Laboratory, Pavia, Italy h CICECO− Aveiro Institute of Materials, Department of Chemistry, University of Aveiro, 3810-193, Aveiro, Portugal A R T I C L E I N F O A B S T R A C T Keywords: Neutron capture therapy (NCT) is a form of radiotherapy that exploits the potential of some specific isotopes to Lithium-6 capture thermal neutrons and subsequently yield high linear energy transfer (LET) particles, suitable for cancer Enriched lithium treatment. Recently, relevant technological improvements have been made in terms of accelerators as suitable 6Li neutron sources for NCT at hospitals. However, low selective delivery of current drugs to cancer cells remains as Carbon nanohorns Short multiwall carbon nanotubes the main challenge for successful clinical application of NCT. This work presents an innovative and previously Cancer unexplored approach for the design of nanotherapeutic NCT agents. Herein, a new concept based on carbon Radiotherapy nanomaterials that seal 6Li active NCT nuclides is investigated. The 6Li active species are located in the inner Nanooncology cavity of the nanocarrier (carbon nanohorns or carbon nanotubes) and therefore, completely protected from the biological environment, avoiding toxicity and degradation. After encapsulation of the active cargo, the external surface of the nanocarrier is modified for improved biocompatibility. The developed 6Li-filled carbon nanohorns offered the possibility to explore 6Li compounds as active NCT agents by delivering therapeutic doses to cancer cells. We envisage that nanoencapsulation of 6Li can trigger the successful development and implementation of Lithium Neutron Cancer Therapy (LiNCT). 1. Introduction recent technological developments that allow the installation of accel­ erators at medical institutions will bring NCT closer to established Cancer is one of the leading diseases worldwide accounting for clinical applications [2,3]. nearly 10 million deaths in 2020 [1]. Neutron capture therapy (NCT) is a Since the early days of NCT, many different delivery agents have highly effective form of hadron-therapy that exploits the potential of been developed and studied [4]. However, only two pharmaceuticals some specific isotopes for cancer treatment, based on the neutron cap­ have been currently employed in clinical trials, namely sodium ture and emission of particles. The short range of the particles, compa­ mercaptoundecahydro-closo-dodecaborate (BSH; Na10 2 B12H11SH) and rable to the average cell diameter, produced in NCT can limit the L-BPA (C9H10 12BNO4) derivatives [5,6]. However, over the years the NCT destructive effects to the cells where the isotopes are localized, thus scientific community has focused on the need of new carriers charac­ conferring high therapeutic precision to this type of radiotherapy. The terized by higher tumour specificity [7–9]. In this respect nanomedicine, * Corresponding author. ** Corresponding author. Institut de Ciència de Materials de Barcelona (ICMAB-CSIC) Campus de la UAB, 08193, Bellaterra, Spain. E-mail addresses: ggoncalves@ua.pt (G. Gonçalves), gerard.tobias@icmab.es (G. Tobías-Rossell). https://doi.org/10.1016/j.carbon.2023.03.034 Received 6 March 2023; Accepted 13 March 2023 Available online 16 March 2023 0008-6223/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). G. Gonçalves et al. Carbon 208 (2023) 148–159 the application of nanotechnology to medicine, has become one of the have been already successfully explored for high resolution bioimaging most promising routes for the targeted diagnosis and treatment of cancer [30–32] and more recently for radioisotope therapy [33–35]. In NCT, [10]. Therefore, novel nanocarriers are currently being explored as de­ the employed isotopes are non-radioactive and only become therapeu­ livery platforms for NCT, in an effort to overcome the limitations of the tically active once they are irradiated with neutrons, reducing in this classical delivery compounds [11–14]. way the risk of secondary effects. In this study, we investigate the Despite there is a wide number of nuclides with high affinity for the nanoencapsulation of enriched 6Li halides into various types of nano­ capture of thermal neutrons, the vast majority of research has been capsules to provide experimental evidence of the potential that the devoted to 10B. Another element that holds great interest for NCT is 6Li. encapsulation of 6Li holds for Lithium Neutron Capture Therapy In fact, the nuclear reactions of 6Li resultant from neutron irradiation (LiNCT). Moreover, to the best of our knowledge, this study provides for produce higher energetic particles than the ones resultant from 10B. the first time biological data using isotopically enriched 6Li for NCT. Therefore the use of 6Li can improve the efficiency of NCT in the erad­ ication of cancer cells. 2. Experimental section The particles generated from its capture reaction are high linear energy transfer (LET) particles, following the reaction depicted below. 2.1. Material synthesis Because of their size and energy, they are confined to a radius of approximately 9–10 μm, comparable to the dimensions of a single cell 2.1.1. Preparation of CNCs from MWCNTs and CNHs [15]. 2.1.1.1. Purification, shortening and tip-opening of MWCNTs. The raw n + 63 Li → 42 He(2.05MeV) + 31 H(2.73MeV) multi-walled carbon nanotubes (MWCNTs) used in this work were ac­ Interestingly, in the early days of NCT, eighty years ago, both 6Li and quired from NanoAmor. The oxidative cutting of the MWCNTs was 10 performed using a combination of acid treatment and ultrasonication, as B were considered as potential candidates for NCT [16–18], but shortly after, boron took the leading role and research with lithium was previously reported [36]. Briefly, 150 mg of MWCNTs were dispersed in discontinued. The reason most likely resides in the different chemistry of 150 mL of a mixture of H2SO4/HNO3 (3/1) and sonicated during 24 h at both elements [15]. Boron can form stable bonds with other elements a temperature of ca. 40 ◦ C. The resulting suspension was washed with such as carbon, oxygen or nitrogen, thus allowing the synthesis of many distilled water and filtered under vacuum using a 0.2 μm PTFE mem­ different compounds, including boron clusters [19,20]. In contrast, brane (Omnipore), until the filtrate reached neutral pH. After that, the lithium is an alkali metal typically encountered as a cation in inorganic MWCNTs were dispersed again in ultrapure water (200 mL) by ultra­ salts and intercalation compounds. Therefore, under biological condi­ sonication, and then centrifuged at either 4000 or 20000 rpm (1503 and tions it would be present as lithium cations. Research with 6Li has thus 37569 rcf respectively) during 30 min. Two different centrifugation been very limited. Rudimentary experiments involving crude neutron speeds were initially employed to assess the degree of removal of large sources were performed back in the 40s [21]. Despite the scientific bundles and long carbon nanotubes. The precipitate was discarded limitations on those early studies, a significant increase in tumour (bundles or long nanotubes), and the process was repeated twice. The regression was observed upon neutron irradiation of a mixture of MWCNTs suspension was collected by vacuum filtration and dried in the lithium and boron (using lithium meta-borate) [16]. In a posterior work, oven overnight at 100 ◦ C. Finally, the prepared short MWCNTs were Cooper et Zahl investigated the energy absorbed by tissue containing reduced by thermal treatment in order to remove oxygen functional either lithium salts or boron. Even though enriched elements were not groups introduced during the shortening step. For this purpose, available at that time, by extrapolation, they could provide evidence of short-MWCNTs (50 mg) were inserted in a silica tube and the sample the potential that 6Li held for NCT [17,18]. Note that naturally occurring was left under vacuum for 2 h. The tube was then placed inside a furnace lithium is composed by only 4.85% of 6Li. and annealed at 800 ◦ C under vacuum during 2 h (heating rate More recently, in 2010 Louis Rendina addressed the following 3.3 ◦ C⋅min− 1). challenge to the scientific community “can lithium salts herald a new era for neutron capture therapy?” [21]. For the past 70 years, lithium has 2.1.1.2. Oxidizing thermal treatment of CNHs for opening the tips. Carbon mostly been utilized to treat bipolar disorder in the clinical setting [22] nanohorns (CNHs) were acquired from Carbonium. The CNHs were and it is on the World Health Organization’s list of essential medicines thermally treated at two different temperatures (in order to open the tips [23]. Herein, we intent to expand the use of lithium and provide evi­ and improve the access to the inner cavity) following the protocol dence that lithium (6Li) can indeed herald a new era for NCT. To over­ described by Utsumi et al. [37]. Briefly, 50 mg of CNHs were placed in a come the limitation that the intrinsic properties of lithium presents to silica boat and then placed in a tubular furnace. The thermal treatment control its biodistribution, lithium salts are sealed in the interior of was carried out under dry synthetic air, reaching a maximum temper­ nanocapsules. The use of nanocapsules offers versatility in terms of ature of 400 or 500 ◦ C (heating rate of 1 ◦ C⋅min− 1). The samples were biodistribution, because it will be insensitive to the innate affinity of the immediately cooled down after reaching the maximum temperature. lithium cation, and it will be merely determined by the physico-chemical properties of the nanocapsules (size, morphology, functional-groups, 2.1.2. Synthesis of 6LiI@CNCs etc.). Our approach also compensates for the lower cross-section of 6Li The filling of MWCNTs was attempted with different lithium com­ vs. 10B because a large amount of 6Li is confined in each nanocapsule. pounds via molten capillarity wetting, following a modified protocol This is a paradigm shift in the strategic development of NCT agents, from the one reported by Sloan et al. for the encapsulation of LiI into since it will prevent the release of lithium cations in biological media. single-walled CNTs [38]. Briefly, 10 mg of shortened MWCNTs were The possibility of encapsulating a high density of 6Li in the form of mixed with 100 mg of the selected lithium compound inside an Ar filled inorganic crystals, plus the fact that the nuclear reaction of 6Li with glovebox (using an agate mortar and pestle), until the sample presented neutrons produces more energetic particles than the ones resulting from homogenous colour. Subsequently, the powder was sealed under vac­ 10 B, can improve the efficiency of NCT in the eradication of cancer cells. uum in a silica tube of ca. 10 cm in length. The silica ampoule was placed We will focus on the use of carbon nanomaterials as 6Li nanocarriers, into a furnace and the mixture was annealed at 5 ◦ C⋅min− 1 until because they are an emerging platform for disease treatment and tissue reaching the first dwell step (50 ◦ C above the melting point of the cor­ repair [24–28]. Carbon nanocapsules (CNCs) based on closed-ended responding salt: LiI, LiCl and LiF acquired from Sigma-Aldrich with filled carbon nanotubes (CNTs), can provide the perfect shelter for the purity >99%). After 4 h, the temperature was increased (5 ◦ C⋅min− 1) up biological stability of high doses of payload [29]. Indeed, filled CNTs to 1100 ◦ C, where the system was dwelled for 10 min. Finally, the 149 G. Gonçalves et al. Carbon 208 (2023) 148–159 sample was cooled down (1 ◦ C⋅min− 1) until a temperature 50 ◦ C below microscope operating at 5 kV. the melting point of the inorganic salt and subsequently the cooling rate Brunauer-Emmett-Teller (BET) analysis was performed on 15–30 mg was increased to 5 ◦ C⋅min− 1 until room temperature. This thermal of raw CNHs after being annealed at 400 ◦ C or 500 ◦ C dry air. The treatment was designed to allow the formation of filled closed-ended samples underwent a treatment of adsorbed gas removal at 150 ◦ C for 2 carbon nanotubes in a single step [39]. For the filling of CNHs with h. This was followed by nitrogen adsorption/desorption. lithium salts, slight variations to the annealing procedure were applied. Thermogravimetric analyses were performed on a Netzsch instru­ The silica ampoule (sealed under vacuum), containing the correspond­ ment, model STA 449 F1 Jupiter®, under flowing air at a heating rate of ing mixture (lithium salt/CNHs) was placed inside the tubular furnace 10 ◦ C⋅min− 1. and then annealed at 950 ◦ C for 12 h. The system was subsequently Samples for X-ray photoelectron spectroscopy (XPS) were prepared cooled down to 600 ◦ C at 1 ◦ C⋅min− 1 and finally to room temperature at by placing dropwise a dispersion prepared by sonicating the sample in 5 ◦ C⋅min− 1. Once the best filling rate of CNCs was obtained for LiI, the dry ethanol onto a silicon wafer (5 mm × 5 mm, previously washed in enriched form of this compound (6LiI) was explored during the NCT water/acetone). XPS measurements were performed at room tempera­ studies. ture with a SPECS PHOIBOS 150 hemispherical analyzer (SPECS GmbH, After the filling step, CNCs were purified by washing the excess of Berlin, Germany) in a base pressure of 5 × 10− 10 mbar using mono­ inorganic material that remained external to the CNCs. The mixture was chromatic Al Kalpha radiation (1486.74 eV) as excitation source. High first dispersed in water (sonication for 15 min) and then refluxed at resolution spectra of C1s, O1s, N1s and I3d regions were also registered. 110 ◦ C during 4 h. This cycle was repeated three times, in order to guaranty the complete removal of the excess of non-encapsulated 2.3. Computer modelling of LiI@CNCs lithium compounds. The purified CNCs were collected by vacuum filtration using a 0.2 μm PTFE Membrane (Omnipore). Afterwards, the Semiempirical quantum chemical calculations were used to explore sample was dried overnight at 100 ◦ C. the molecular structure and energetics of LiI adsorbed and encapsulated in nanocarbon materials. The xtb-GFN1 method [41], previously used to 2.1.3. Synthesis of EuCl3@CNCs explain graphene-inorganic interfaces [42], was used within the peri­ The sample of EuCl3@CNHs was prepared following a similar pro­ odic DFTB+ 22.1 software [43]. The LiI@graphene and LiI@CNT in­ tocol to that described for the encapsulation of lithium halides into terfaces were built and modified with the aid of Virtual NanoLab [44] CNHs. For that purpose, 15 mg of oxidized CNHs were mixed with 200 and the Atomic Simulation Environment (ASE) [45]. A (38,0) semi­ mg of EuCl3 and then sealed in a silica tube under an inert atmosphere. conducting single wall carbon nanotube was selected to represent LiI The silica ampoule was then placed into a furnace, and the mixture was encapsulation. To compute the relative Li at. %, semiconducting annealed at 5 ◦ C⋅min− 1 until reaching 950 ◦ C. After 12 h, the temper­ MWCNTs were built by repeatedly adding CNTs with a 3.52 Å larger ature was lowered to room temperature at a rate of 10 ◦ C⋅min− 1. After radius [46]. 3D periodic cells with 50 nm and 10 nm perpendicular axis the filling step, CNCs were purified by washing the excess of inorganic to the molecular directions, i.e. layer and axis, were used to simulate the material that remained external to the CNCs, as previously described. 2D and 1D systems respectively. The periodic cells used fixed slab and axis dimensions of 3.33 nm and 3 nm for graphene and the CNT 2.1.4. External surface functionalization of 6LiI@CNCs respectively and were computed at the Γ point. The external surface functionalization of the CNCs with free amine groups was performed using diazonium-based arylation reaction [40]. 2.4. Internalization studies by confocal microscopy First, the CNCs (20 mg) were dispersed in DMF (20 mL) by ultra­ sonication during 30 min. After that, 200 mg of 4-[(N-Boc)amino­ To evaluate CNHs internalization in HeLa cells, EuCl3@CNH were methyl]aniline were added to the dispersion and dissolved by employed because europium allows their direct detection in confocal mechanical stirring. The mixture was cooled down to 0 ◦ C and then 200 microscopy by light reflection. HeLa cells were selected as they repre­ μL of isopentyl nitrite were added. The reaction temperature was raised sent a highly used in vitro model that it is proven to be useful as a starting to 80 ◦ C and kept stirring during 4 h. The sample purification was per­ point for research. A large amount of information is available about their formed by cycles of washing and filtering under vacuum using the properties and growing conditions. following sequence of solvents: DMF/water/methanol/diethyl ether. To prepare the samples, sterile 20 mm × 20 mm coverslips were put Finally, the sample was dried under vacuum overnight. inside 35 mm dish and 3 × 104 cells were seeded on them. EuCl3@CNHs The deprotection of Boc group at the surface of the CNCs was per­ were sterilized under UV light for 1 h and then dispersed in culture formed by treatment with a 4 M HCl solution in dioxane. The resulting medium and ultrasonicated obtaining a concentration of 0.02 mg⋅mL-1. CNCs were initially well dispersed in 20 mL of dioxane solution by Cells were allowed to grow for 24 h at 37 ◦ C and 5% of CO2, and then 1 ultrasonication during 15 min and then, the dispersion was kept stirring mL of EuCl3@CNHs solution was added to each dish without removing overnight at room temperature. The sample purification was performed the previous millilitre obtaining a final concentration of 0.01 mg⋅mL-1. by vacuum filtration using the following sequence of solvents: water/ Cells were incubated with CNHs for 24 h; after the contact time, medium methanol/diethyl ether. Finally, the samples were dried under vacuum was removed and cells were washed with DPBS (Dulbecco’s phosphate overnight. buffered saline: 200 mg⋅mL-1 KCl, 200 mg⋅mL-1 KH2PO4, 8000 mg⋅mL-1 NaCl and 2160 mg⋅mL-1 Na2HPO4–7H2O; ThermoFisher Scientific). 2.2. Characterization Then, cells were fixed with paraformaldehyde (PFA) 4% for 20 min and washed with DPBS. Once they were fixed, HeLa cells were stained with Samples for microscopic characterization were prepared by soni­ Hoechst 33342 10 mg⋅mL-1 (InvitrogenTM), which stains the nuclear cating a small amount of CNCs in absolute ethanol. The sample was then DNA, and Cell Mask Deep Red Membrane 5 mg⋅mL-1 (InvitrogenTM), placed dropwise onto a holey carbon grid. Transmission electron mi­ which stains the plasma membrane. Dye dilutions in DPBS (1:1000) croscopy (TEM) images were obtained using a JEOL 1210 microscope, were prepared and 1 mL of them was added to the cells, they were operating at 120 kV. High resolution TEM (HRTEM), high angle annular incubated for 10 min at room temperature and shielded from light dark field scanning transmission electron microscopy (HAADF STEM) before being washed with DPBS, staining first the membrane and then and energy dispersive X-ray spectroscopy were carried out in a FEI the nucleus. After the staining, coverslips were transferred to glass slide Tecnai G2 F20 microscope operated at 200 kV coupled to an EDAX super using mounting media prolong glass to attach them; they were kept in ultra-thin window (SUTW) X-ray detector. Scanning electron micro­ the fridge until they were analysed with the confocal microscope. scopy (SEM) images were acquired using a Quanta FEI 200 ESEM FEG For the acquisition, Leica TCS SP5 (Servei de Microscòpia de la UAB) 150 G. Gonçalves et al. Carbon 208 (2023) 148–159 was employed using HCX PL APO lambda blue 63x/1.4 OIL UV objec­ pellets with a second irradiation. KBr pellets were placed on mylar disks tive. Lasers used for image acquisition were HeNe 633 nm (10%) for Cell and irradiated in contact with the CR39 films in the same way as cell Mask Deep Red Membrane (detection range from 643 nm to 782 nm), pellets. Cell samples were placed on another CR39 and irradiated in the Diode UV 405 nm (6%) for Hoechst 33342 (detection range from 413 nm same position for 2 h at 250 kW. The higher neutron fluence causes a to 478 nm) and Argon laser at 488 nm (20%) for europium reflection higher track density on the detector, which constitutes a map of Li dis­ (detection range from 483 nm to 497 nm). Final images were processed tribution in the samples [49]. Qualitative autoradiography results in a using ImageJ software. black and white picture, where the gray halo represents the region with lithium uptake. The whiter parts of the images are the ones with higher 2.5. Neutron irradiation experiments lithium concentration. 2.5.1. Sample preparation for neutron irradiation 2.5.4. In vitro neutron irradiation experiments Neutron irradiation experiments with 6LiI@CNCs were performed NCT effects were studied using the same cell line described above. To using KBr pellets. The pellets were prepared by grinding 200 mg of KBr evaluate the effect of neutron capture in lithium, four flasks were pre­ with 2 mg of 6LiI@CNCs using an agate mortar and pestle. The resulting pared: 2 serving as the control, without any treatment (non-enriched mixture was pressed (5 tons) using a mould pellet press die with a cells) and 2 where cells were exposed to the 6LiI@CNCs (enriched cells). diameter of 1.0 cm. These pellets were prepared to study the presence Rat osteosarcoma UMR-106 cells were cultured in medium composed by and the uniformity of lithium in nanomaterials, by means of qualitative DMEM high glucose, 10% FCS and 1% gentamicine, then seeded at the autoradiography (see below). density of 3 × 106 cells in four 75 cm2 flasks. After 48 h, for two flasks, the culture medium was replaced and cells were allowed to grow for 4 h 2.5.2. Cell preparation and treatment with 6LiI@CNCs in the medium enriched with the carbon nanomaterials containing To evaluate the uptake, a rat osteosarcoma UMR-106 cell line was lithium, following the same procedure as described in Section 2.2.2. At employed, as extensive experience of BNCT has been acquired using this the end of the contact time, cells were washed, and fresh medium was model [47]. Cells were cultured in medium composed by DMEM high added to the flask. The same procedure was performed for the other two glucose, 10% FCS and 1% gentamicine. Then seeded at the density of 3 control flasks. Finally, a non-enriched and an enriched cell flask were × 106 cells in two 75 cm2 flasks. After 48 h, the culture medium was irradiated in the thermal column of the TRIGA Mark II reactor for 11 min replaced, and cells were allowed to grow for 4 h in the medium enriched at 250 kW. The irradiation position is characterized by a thermal with the 6LiI@CNCs. The 6LiI@CNCs contained in each vial (4 mg or 0.4 neutron flux of about 1010 cm− 2 s− 1 at 250 kW. Cell survival was eval­ mg), were dispersed in 20 mL of culture medium hand shaked until uated by means of cloning assay, measuring the plating efficiency. dissolved and then ultrasonicated for 30 min, leading to final concen­ trations of 0.2 and 0.02 mg⋅mL-1. Before adding the enriched cell culture 3. Results and discussion medium to the cell culture flasks, the enriched medium was exposed to UV light for 15 min to avoid cell contamination. At the end of the fixed 3.1. Preparation of CNCs from MWCNTs and CNHs contact time, after the medium removal, the cells were washed five times in DPBS (Euroclone Cat n◦ ECB4004L), trypsinized and counted. Within this work, two different types of carbon nanostructures, Finally, four million cells were selected and centrifuged, forming a namely MWCNTs (Fig. 1a) and CNHs (Fig. 1e) were explored as nano­ condensed cell pellet. The pellet was then deposited on a mylar disk and carriers for the delivery of enriched 6Li into cancer cells. We notice that left drying. the preparation of viable carbon nanomaterials for melt filling is dependent on the employed carbon nanostructure. In the case of 2.5.3. Neutron autoradiography MWCNTs, a two-step treatment was usually needed for the preparation Neutron autoradiography is based on the use of passive detectors of nanocapsules. The first step consists in an oxidizing protocol, which sensitive to the charged particles radiation. In this work, we used rect­ combines a mixture of H2SO4/HNO3 and ultrasonication. Acid treatment angular polyallyldiglycol carbonate (PADC) CR39 films. Samples con­ is a simple and efficient approach that allows cutting MWCNTs, removes taining elements capturing neutrons and emitting charged particles are metal particles (catalyst employed for the growth of CNTs) and induces irradiated in contact with the detector. After irradiation, the detector is the formation of oxygen-containing functional groups along the CNTs etched in a chemical solution that enlarges the latent tracks and makes surface [36]. With the aim to discard long CNTs or bundles that can still them visible for microscopical analysis. Depending on the calibration of be present after the oxidative cutting, we took advantage of the water the technique (neutron fluence, etching parameters), it can be employed dispersibility of the acid treated MWCNTs to fractionate them by length to obtain quantitative information on the element concentration in via centrifugation (at 4000 and 20000 rpm). The precipitated material at sample or qualitative imaging of its spatial distribution. the bottom of the centrifuge tube was discarded and the supernatant was Quantitative neutron autoradiography was used to check the lithium collected. In this way, we could obtain purified open-ended and very uptake in cell samples and the presence of lithium in KBr pellets, by short MWCNTs with a narrow length distribution (Fig. 1b). detecting α tracks produced in the thermal neutron capture reaction on Fig. S1 shows the length distribution of the acid cut MWCNTs after 6 Li. applying different centrifugation speeds, namely, 4000 and 20000 rpm. The dry cell samples on the mylar disks were irradiated two times in The length distribution was determined by measuring over 1000 spec­ contact with CR39 in the thermal neutron column of the TRIGA MARK II imens from SEM images. Statistical analysis reveals an average length nuclear reactor at Laboratorio Energia Nucleare Applicata (LENA) of the distribution of 186.2 ± 81.9 and 144.3 ± 53.9 nm for the MWCNTs University of Pavia. For the first irradiation, samples were irradiated for centrifuged at 4000 rpm (Figure S1 a) and b)) and 20000 rpm (Figure S1 30 min at 2 kW reactor power, receiving a thermal neutron fluence of 1.9 c) and d)), respectively. Therefore, due to the small average size and ± 0.1 × 1010 cm− 2. This neutron autoradiography set-up can provide narrow distribution, MWCNTs obtained at the highest centrifugation quantitative autoradiography analysis by counting tracks generated by speed were employed in subsequent studies. It is important to notice that the α particles in the CR39 detector using a LEICA MZ16A microscope the obtained length distributions are within the range of improved [48]. The results are presented as a composition of 40 (8 × 5) sequential biocompatibility for CNTs [50]. AFM imaging corroborates the very pictures. Each image has an area of 0.3 mm2 (0.632 mm × 0.474 mm), short length of MWCNTs observed by SEM and give an additional giving a total measured area of 12 mm2, which is a representative overview of the morphology of the samples (Fig. S2). Subsequently, portion of the entire sample. short MWCNTs were thermally reduced in order to remove the Qualitative images were obtained both for cell samples and for KBr oxygen-based functionalities introduced on their graphitic structure by 151 G. Gonçalves et al. Carbon 208 (2023) 148–159 Fig. 1. Synthesis and characterization of carbon nanocapsules, short-MWCNTs and CNHs, filled with 6LiI. a) Schematic representation of the synthesis of 6 LiI@MWCNTs. b) TEM image of the short-MWCNTs before filling with red arrows indicating the open tips; the inset shows the size distribution determined by SEM. c) HAADF STEM image of 6LiI@MWCNTs showing high contrast of I (corresponding to 6LiI) in the inner cavity of short-MWCNTs. d) HR-XPS spectrum of I3d corresponding to 6LiI@MWCNTs. e) Schematic representation of the synthesis of 6LiI@CNHs. f) TEM image of the CNHs before filling, with the inset showing the diameter distribution determined by SEM. g) HAADF STEM image of the 6LiI@MWCNTs showing homogeneously distributed high contrast of I (from 6LiI) in the inner cavities of CNHs. h) HR-XPS spectrum of I3d from 6LiI@MWCNTs. the acid treatment. The elimination of the aliphatic fractions was annealed at 400 ◦ C (CNHs400◦ C = 1317 m2⋅g− 1) and 500 ◦ C (CNHs500◦ C confirmed by thermogravimetric analysis (TGA, Fig. S3). Whereas the = 1857 m2⋅g− 1) reveals the impact that temperature has on the degree of TGA curve of the acid treated MWCNTs shows a continuous weight loss tip opening (Fig. S4). The largest surface area is reached at the highest due to the removal of adsorbed water and functional groups (red line), employed temperature. The value of SA achieved for CNHs500◦ C is the vacuum annealed sample becomes thermally stable until its com­ significantly higher than that reported by Fan et al. (1450 m2⋅g− 1). It plete combustion at ca. 550 ◦ C (green line). almost reaches the value achieved for graphene [51], showing that both The preparation CNHs suitable for the synthesis of CNCs is simpler sides of the single carbon layer in CNHs are fully accessible [37]. and only requires a single oxidizing step for the controlled tip hole Additional characterization using TGA, under flowing air, reveals that opening of their carbon structures (Fig. 1e) [37]. This controlled thermal both thermally treated CNHs present a similar combustion curve to oxidation can be applied because disruption of CNHs preferentially oc­ as-received CNHs (Fig. S5). This indicates that the controlled oxidation curs through the tips, where a high density of structural defects or high at 400 ◦ C and 500 ◦ C takes only place in the tips and preserves the strain is commonly present. As received CNHs were independently structural integrity of CNHs. Consequently, the CNHs with highest SA treated at 400 ◦ C and 500 ◦ C in air. The collected samples were analysed were selected for the development of nanocapsules. By having a large by BET to determine the specific surface area (SA), which allows esti­ number of opened tips, it is expected to maximize the filling rate with mating the degree of tip opening at each treatment temperature. Com­ the lithium compounds. parison of the SA of raw CNHs (CNHs = 419 m2⋅g− 1) and after being Fig. 1f shows a TEM image of CNHs after the oxidizing thermal 152 G. Gonçalves et al. Carbon 208 (2023) 148–159 treatment at 500 ◦ C. Statistical analysis reveals a median diameter of sample turns out to be 1.4 at. % for 6LiI@MWCNTs and 3.5 at. % for 6 59.3 ± 16.9 nm (inset Fig. 1f) determined by measuring over 500 LiI@CNHs (Table S1), corroborating the higher loading of the CNHs specimens from SEM images. AFM and SEM images of CNHs are shown cavities with respect to MWCNTs already shown by EDX. in Fig. S2 and Fig. S6. The successful formation of hermetically sealed nanostructures with 6 LiI nanometric crystals in the interior of MWCNTs and CNHs was 3.2. Encapsulation of 6Li onto CNCs (6LiI@CNCs) confirmed by HRTEM (Fig. 2). Furthermore, it is also possible to confirm through visual inspection that the graphitic ends of the MWCNTs and To maximize the lithium loading in short MWCNTs, via molten phase CNHs (yellow arrowed) are reconstructed, forming in this way the capillary wetting, several lithium salts were attempted, namely lithium desired hermetically sealed carbon nanocapsules (see Fig. S12 for iodide, lithium chloride and lithium fluoride. These salts were inde­ additional images). Through intensity profiles acquired on different pendently ground with a given amount of MWCNTs and sealed under crystals we could determine interplanar spacings of 3.5 Å and 3.0 Å. vacuum in a silica ampoule. The mixture was then heated above the These are in good agreement with the (111) plane and (200) of cubic LiI melting temperature of the salt to allow capillary wetting of the inner (Fm-3m (225), ICSD), respectively, and computer models of encapsu­ cavities of the open-ended MWCNTs by the molten halide. The obtained lated LiI (see below). results showed that the high reactivity of lithium chloride and lithium fluoride with the silica ampoules at the temperatures employed for the 3.2.1. Computer modelling of LiI@CNCs filling step restrict their use in this approach (Fig. S7). Consequently, To get insight into the filling and interaction of LiI with the graphene only lithium iodide (6LiI) was explored on this work for the preparation layers of the CNCs, computer models of LiI encapsulated into and onto of CNCs. In general, bulk filling of carbon nanomaterials results in the graphene sheets and inside CNTs were studied. First, we investigated the presence of a large amount of non-encapsulated compounds [52]. interface of an increased number of LiI (100) monolayers ranging from Therefore, after the filling experiment, 6LiI that remained external to the one to six on a graphene square surface 3.3 × 3.3 nm containing 418 MWCNTs or CNHs was removed from the sample by extensive washings carbon atoms. The binding energy of LiI on graphene increases mono­ in water (LiI solubility = 1.67 g⋅mL− 1 at 25 ◦ C). tonically with the number of layers and is mostly converged from four The successful filling of MWCNTs and CNHs with 6LiI was initially LiI layers with a binding energy of 28 meV per carbon atom (Fig. S13). determined by Z-contrast imaging. Imaging of the encapsulated com­ This is equivalent to an energy for the graphene-LiI interface of 0.17 J⋅m- 2 pound by Z-contrast STEM was possible because a large contrast is , which is roughly half the energy binding two graphene layers observed between I (Z = 53) and C (Z = 6) as a result of the Rutherford computed with the same model (0.30 J⋅m-2). Next, to understand better scattering used to form Z-contrast images, which scales approximately encapsulation, we computed systems with four, five and six LiI mono­ with the square of the atomic number [53,54]. This imaging modality layers between two graphene layers. These monolayers were found to be has been widely employed to determine the crystal structure of inor­ 2.3, 2.0 and 1.7 nm apart, respectively, from which a LiI interlayer ganic compounds [55,56]. In the presence case, iodide atoms, from 6LiI, distance of 3.00 Å can be easily derived, in good agreement with the are expected to appear much brighter than carbon. On the other hand, experimental measurements. lithium, with a Z = 3, cannot be observed in this imaging modality. The In addition, an analysis of the binding energies of both systems presence of bright fragments inside the darker carbon shells confirms the (Fig. 3) reveals that the binding between LiI adsorbed and encapsulated successful loading of 6LiI in the inner cavities of MWCNTs and CNHs on layered Csp2 materials are virtually equivalent for the studied (Fig. 1c and g, respectively, and Fig. S8). Moreover, bright particles were systems. not observed external to the walls of the CNCs, confirming the full The molecular structure and energetics of LiI encapsulated in a 3 nm removal of the excess of 6LiI employed in the synthesis. The selective wide CNT containing 1064 carbon atoms was also studied. Two systems dissolution of free 6LiI while preserving the encapsulated material, of the were computed, namely a LiI nanowire and a monolayer of LiI coating same nature, indicates that the ends/tips of the carbon nanomaterials the inner CNT wall (Fig. 3 and Fig. S14), yielding binding energies of 21 became closed during the filling experiment. This hermetically sealed meV and 15 meV per carbon atom, respectively. This means that, if carbon system isolates the encapsulated compound from the external needed, the release of Li would be energetically more likely for the medium, which is an essential aspect for the development of CNCs for monolayered LiI than for fully filled-up systems. The relative atomic biomedical purposes. It has been previously reported that the ends of percentages of Li are 16 at. % and 9 at. % for the LiI nanowire and LiI MWCNTs can be sealed by thermal annealing, provided when a high monolayer respectively. However, these values depend on the number of enough temperature is given to the system [39]. Here we also provide CNT walls, and for instance, for a MWCNT with six walls, the atomic evidence that the tips of CNHs are also able to be regenerated by percentages of Li get reduced to 2 at. % and 1 at. % for the LiI nanowire adequate thermal treatment. Thermal regeneration of the ends of carbon and the monolayer of LiI respectively, which are in the range of the nanostructures depends on their diameter. A higher temperature is for experimental quantification. instance needed to seal multi-walled CNTs compared to their single-walled counterparts [39,57,58]. In the case of short-MWCNTs we 3.3. Cell internalization studies employed 1100 ◦ C, whereas in the case of CNHs we used 950 ◦ C to promote simultaneously the process of filling and end closing. Cell internalization of CNCs is necessary for their application in 6 Energy dispersive X-ray analysis (EDX) was next carried out to LiNCT. Therefore, after having proved that both CNTs and CNHs can be confirm the presence of I and for the semi-quantitative determination of successfully filled, cell internalization studies were performed using the atomic composition of the synthesized materials. An estimated CNHs. We focused on CNHs because filled CNTs have been extensively concentration of 0.25 at. % I and 0.63 at. % I was determined for studied for biomedical imaging and therapy [59]. For this purpose, LiI 6 LiI@MWCNTs and 6LiI@CNHs, respectively (Figs. S9–S10) indicating a was replaced by EuCl3 using exactly the same protocol for the prepa­ higher metal halide loading in CNHs than in MWCNTs. X-ray photo­ ration of the nanocapsules. EuCl3@CNHs were used to evaluate inter­ electron spectroscopy (XPS) data also confirm the presence of I (from nalization since europium can be visualized by confocal microscopy. 6 LiI) filled into both MWCNTs and CNHs cavities, with signals at ca. 620 HeLa cells were exposed to 0.01 mg⋅mL-1 of EuCl3@CNHs for 24 h. Then, eV corresponding to I3d (Fig. S11). The signal intensity of lithium was they were fixed and dyed with Cell Mask Deep Red Membrane to stain too low to be measurable by this technique. Therefore, the concentration the cell membrane and Hoechst 33342 to stain the nucleus, allowing the of enriched lithium was determined based on the quantification of definition of cellular compartment and the better interpretation of iodine in the I3d5/2 high-resolution spectrum (Table S1). Considering particles location. Due to the fixation of the cells before the staining, Cell that I and Li are equimolar (1:1 ratio), the atomic percentage of 6Li in the Mask Deep Red Membrane dyed not only the membrane but also the 153 G. Gonçalves et al. Carbon 208 (2023) 148–159 Fig. 2. Electron microscopy characterization of carbon nanocapsules. HRTEM images of a, b) 6LiI@MWCNTs and c, d) 6LiI@CNHs. Lattice fringes can be observed for several crystals of 6LiI. As guide to the eye, white arrows point to encapsulated 6LiI, and yellow arrows to closed ends. Fig. 3. Quantum chemical models of LiI@CNT, namely: a) six monolayers of LiI adsorbed on graphene, b) six monolayers of LiI encapsulated between two graphene sheets, c) LiI nanowire filling up a 3 nm wide CNT and d) LiI monolayer in a 3 nm wide CNT. Fig. 4. Orthogonal projections of confocal images of HeLa cells incubated with 0.01 mg⋅mL-1 EuCl3@CNHs in medium for 24 h. White lines determine the cut cytoplasm of the cells. Confocal images revealed the presence of the in each plane and the white circle indicates the position of some of the inter­ EuCl3@CNHs (in green) inside HeLa cells. To precisely determine the nalized EuCl3@CNHs. EuCl3@CNHs are shown in green, nucleus in blue and location of EuCl3@CNHs in the cells, orthogonal projections were per­ cytoplasm in red. formed. This allows the localization of the compounds in the three di­ mensions (x, y, z). Fig. 4 shows the orthogonal projections of two cells. 154 G. Gonçalves et al. Carbon 208 (2023) 148–159 autoradiography analysis resulted in 44 ± 8 tracks mm− 2, where the uncertainty is the standard deviation of the tracks computed over the 40 images. These analyses confirm the presence of active lithium species in the sample. 3.5. External surface functionalization of CNCs (NH2_6LiI@CNCs) The external surface functionalization of the 6LiI@CNCs was per­ formed using diazonium-based arylation reaction in presence of [(N- Boc)aminomethyl]aniline [40]. Fig. 6 a) shows a schematic represen­ tation of the included functional groups. After deprotection, the pres­ ence of ammonium groups on the external aromatic structure of the nanocarbon shelter confers good dispersibility in water and cell culture medium (Figs. S15 and S16) and it is expected to enhance its biocompatibility. The composition of 6LiI@CNC after chemical modification was determined by XPS analysis (Fig. 6 c and d). High resolution XPS spectra of the C1s, N1s and I3d were employed to quantitatively assess the concentration of C, N and I in the benzylamino functionalized samples. After considering the presence and stoichiometric ratio of both Li (from LiI) and H (present in the benzylamonium group), analyses reveal a similar relative atomic content of N1s for both 6LiI@CNCs (1.9 at. %). The degree of functionalization was calculated taking into account the slights variations in C and I (the latter being higher in the case of NH2_6LiI@CNHs (1.0 at. % of I) than NH2_6LiI@MWCNTs (0.7 at. % of I)), the presence of H (7 H atoms per benzylamonium group containing moiety) and the equimolar concentration of I and 6Li. This led to a surface chemical functionalization of 1.61 and 1.67 mmol⋅g− 1 of CNHs Fig. 5. a) Qualitative neutron autoradiography of the KBr pellet containing and MWCNTs, respectively. These results indicate that the yield of the 6 LiI@MWCNTs. b). Quantitative distribution of lithium in a portion of arylation for the covalent surface modification of the employed carbon the sample. nanostructures is barely affected by their morphology. As a guide to the eye, some of the internalized EuCl3@CNHs are marked 3.6. Biocompatibility and neutron autoradiography with a white circle in each plane, showing their location inside the cytoplasm of HeLa cells. This analysis confirms internalization of the Cytotoxicity studies of NH2_6LiI@CNCs were performed using UMR- CNHs into the cells even at the low studied concentration of 0.01 106 cells, a well-known cell line in BNCT. Results showed that for both mg⋅mL-1. concentrations tested, namely 0.02 and 0.2 mg⋅mL-1, the NH2_6LiI@MWCNT presented a lower cell viability, of about 75%, when compared with NH2_6LiI@CNHs that is >90% (Fig. 7a). The higher 3.4. Quantitative neutron autoradiography cytotoxicity of NH2_6LiI@MWCNT can be attributed to the morphology of the CNCs in this case, because cells are not in contact with the To test whether the developed 6LiI@CNCs held potential for LiNCT, encapsulated 6LiI, and the external functional groups are the same in an initial assessment was performed by neutron autoradiography using both types of carbon nanostructures. Other groups have also compared 6 LiI@MWCNTs dispersed in KBr pellets. 6LiI@MWCNTs were chosen for these two types of nanocarbons. For instance, He et al. recently reported these initial measurements because they have a lower content of 6Li than an improved biocompatibility of CNHs over four different types of CNTs 6 LiI@CNHs. Therefore, if 6Li nuclei reacting with neutrons are already in macrophages [60]. In fact, it is well known that the physico-chemical detected, the same would be expected for 6LiI@CNHs. Fig. 5 shows the properties of nanomaterials, which can be tuned during their synthesis, imaging of lithium distribution in KBr pellets. In the upper part, the determine their biological behaviour [61]. qualitative neutron autoradiography is reported. This analysis allows As the cell toxicity was acceptable, at least for one of the two for­ pointing out the presence of neutron-capture elements and to evaluate mulations proposed, the next step consisted in studying the uptake of their distribution over the surface of the detector. In fact, it is possible to lithium in cells. When a cell is irradiated after the absorption of lithium, see a distribution of clear structures over the black background. These the neutron capture reaction releases charged particles that are structures are the tracks left by charged particles produced by neutron responsible of the inactivation of the tumour cell itself, as shown in capture in lithium, enlarged by means of chemical etching and illumi­ Fig. 7b. This same reaction is exploited by neutron autoradiography. The nated by a microscopy light field. In Fig. 5b, the quantitative analysis of cell pellet previously treated with lithium-containing nanocapsules were a portion of the upper imaging (Fig. 5a) has been reported. Quantitative irradiated in contact with the detectors, again applying the protocols of analysis refers to the count of tracks in single frames taken with the qualitative and quantitative autoradiography. Fig. 8a and b shows the microscope in the same sample but irradiated with lower fluence to qualitative neutron autoradiography of the 6Li@CNC respectively pro­ obtain a track distribution without superpositions. Another CR39 film is duced by MWCNTs and CNHs, where the distribution of nanotubes is employed for this analysis. Each box in the zoomed part of the figure again represented by the white structures visible over the black back­ reports the number of tracks counted in the frames. It is important to ground. The image in Fig. 8a shows that the distribution of nanoparticle notice that tracks are distributed over the entire pellet. Since MWCNTs produced via MWCNT inside cells is not uniform: they are concentrated tend to aggregate in bundles and the KBr pellet was prepared by simple in fact in two areas, while the large majority of the cell pellets does not grinding, the distribution of tracks is inhomogeneous. The in­ evidence presence of lithium. This inhomogeneity may reflect a poor homogeneities are shown by the white dots, which are track clusters dispersibility of the compound in the cell culture medium and conse­ caused by aggregates of lithium atoms. The quantitative neutron quently a limited or no availability of lithium for the majority of the cells 155 G. Gonçalves et al. Carbon 208 (2023) 148–159 Fig. 6. a) Schematic representation of the chemical surface modification of 6LiI@CNHs using diazonium-based arylation reaction in presence of [(N-Boc) aminomethyl]aniline. b, c) High-resolution XPS spectra over C1s, N1s and I3d for functionalized CNCs. in the flask. This result suggests that this production route is not optimal efficiency after neutron irradiation; results are shown in Table 1. When for LiNCT: if only part of the tumour cells absorbs lithium, neutron the cells are not irradiated but are enriched with lithium, the cell sur­ irradiation would generate a non-uniform dose deposition and portion vival is 92%, confirming the preliminary results obtained in the of tumour may survive. On the other hand, lithium enriched nano­ described experiments of biocompatibility. When cells are irradiated particles produced by CNH absorb lithium more homogeneously, as without lithium treatment, the survival drops at 4.2%, mainly due to demonstrated by Fig. 8b. In this image, the white structures are evenly neutron interaction with hydrogen and nitrogen in cells (background distributed on the surface of the pellet, corresponding to a uniform dose). When cells are irradiated after treatment with NH2_6LiI@CNHs, lithium concentration in cells. This is also confirmed by the quantitative the survival is halved (1.8%) due to effect of neutron interaction with analysis, resulting in an average track density of 23 ± 5 mm− 2. The lithium. Being an initial test, a high neutron fluence was employed to uncertainty associated to this measure is the poissonian error of the increase the chances of observing an effect. Under the present dose counted tracks, reflecting the track distribution in the sample: a low conditions, it was observed a cell survival of 4.2%, that is consistent with relative uncertainty means that track density is very regular and does previously reported results [47]. For lithium-treated cells, irradiated not fluctuate much between one frame (0.3 mm2 image) to another. with the same fluence, a remarkable reduction in cell survival of more Therefore, the 6LiI@CNCs based on CNHs hold higher potential for than 50% was achieved when compared with untreated ones (Table 1). LiNCT as they would guarantee the deposition of a uniform dose in the With the present approach the encapsulated 6Li remains inactive tumour. See Figs. S17 and S18 for additional images. until it is activated by an external neutron irradiation. This allows for the nanocapsules to be internalized by cells or even to be delivered to the 3.7. In vitro neutron irradiation experiments diseased tissue in vivo with no biological effect (the latter being outside the scope of the present study). Only when the nanocapsules are located The NH2_6LiI@CNCs were used for a preliminary measurement of into the targeted site, they would be irradiated with external neutrons cell (Rat osteosarcoma UMR-106 cells) survival curves with plating and emit lethal products, thus behaving as carbon nuclear nanobombs 156 G. Gonçalves et al. Carbon 208 (2023) 148–159 Fig. 8. Qualitative neutron autoradiography of a cell pellet exposed to a) NH2_6LiI@ MWCNTs and to b) NH2_6LiI@ CNHs at a concentration of 0.2 mg⋅mL-1. Table 1 Fig. 7. Biocompatibility and neutron autoradiography studies of the developed Cell survival results obtained by plating efficiency. nanocapsules with UMR-106 cells. a) Cytotoxicity studies of NH2_6LiI@CNCs with UMR-106 cells. b) Schematic representation of neutron irradiation of Li enriched Irradiated Cell Survival (%) NH2_6LiI@CNCs inside of cell, and respective formation of nuclear species that can promote cell death. (CNBs). Next steps for the implementation of LiNCT include in vitro neutron dose optimization, a higher encapsulation efficiency of 6Li and the use of targeted nanocapsules for the selective accumulation of 6Li at the tumor site. 4. Conclusions irradiation without 6Li. In addition, the NH2_6LiI@CNHs demonstrated We have successfully developed new CNCs for the containment of low toxicity at the cellular level in absence of irradiation. These results 6 LiI using both CNHs and MWCNTs. With neutron autoradiography we give a first indication of the possibility of using these CNCs for LiNCT have shown that 6Li can be efficiently internalized by UMR cells. The and encourage us to plan a systematic study to estimate the dose in cells accumulation of 6Li in the cells is more homogeneous when using due to the reaction, with the aim to produce a complete cell survival NH2_6LiI@CNCs produced with CNHs compared to that produced with curve. This will allow to evaluate the effective therapeutic potential of MWCNTs. In fact, in the first case the concentration of 6Li present in the the developed lithium-filled nanocapsules. The present study expands cells produced an average spatial track density of 23 ± 5 mm− 2 in the the potentialities that LiNCT holds for the treatment of cancer. CR39, while in the second case this was very inhomogeneous. In addi­ tion, the NH2_6LiI@CNHs demonstrated low toxicity at the cellular level. A first preliminary test of thermal neutron irradiation of UMR cells Declaration of competing interest treated with NH2_6LiI@CNHs showed that, with the concentration of 0.2 mg⋅mL-1, the survival of the cells is reduced compared to an The authors declare the following financial interests/personal 157 G. Gonçalves et al. Carbon 208 (2023) 148–159 relationships which may be considered as potential competing interests: [11] K. Hu, Z. Yang, L. Zhang, L. Xie, L. Wang, H. Xu, L. Josephson, S.H. Liang, M.- R. Zhang, Boron agents for neutron capture therapy, Coord. Chem. Rev. 405 priority patent application entitled "Lithium filled nanocapsules and use (2020), 213139, https://doi.org/10.1016/j.ccr.2019.213139. thereof", P202230271. [12] D. Alberti, A. Deagostino, A. Toppino, N. Protti, S. Bortolussi, S. Altieri, S. Aime, S. Geninatti Crich, An innovative therapeutic approach for malignant mesothelioma treatment based on the use of Gd/boron multimodal probes for MRI Acknowledgements guided BNCT, J. Control. Release Off. J. Control. Release Soc. 280 (2018) 31–38, https://doi.org/10.1016/j.jconrel.2018.04.043. G. T. acknowledges funding from ERC Consolidator Grant NEST [13] K.R. Pulagam, M. Henriksen-Lacey, K.B. Uribe, C. Renero-Lecuna, J. Kumar, A. Charalampopoulou, A. Facoetti, N. Protti, V. Gómez-Vallejo, Z. Baz, V. Kumar, (725743). G.G. gratefully acknowledges the funding by the Portuguese A. Sánchez-Iglesias, S. Altieri, U. Cossío, D. Di Silvio, A.M. Martínez-Villacorta, Science Foundation (FCT) for Programme Stimulus of Scientific A. Ruiz de Angulo, L. Rejc, L.M. Liz-Marzán, J. Llop, In vivo evaluation of Employment – Individual Support (CEECIND/01913/2017), financial multifunctional gold nanorods for boron neutron capture and photothermal therapies, ACS Appl. Mater. Interfaces 13 (2021) 49589–49601, https://doi.org/ support of project CARBONCT (2022.03596.PTDC), TEMA UIDB/ 10.1021/acsami.0c17575. 00481/2020 and UIDP/00481/2020; and CENTRO-01-0145-FEDER- [14] T.V. Popova, I.A. Pyshnaya, O.D. Zakharova, A.E. Akulov, O.B. Shevelev, 022083 - Centro Portugal Regional Operational Programme (Cen­ J. Poletaeva, E.L. Zavjalov, V.N. Silnikov, E.I. Ryabchikova, T.S. Godovikova, Rational design of albumin theranostic conjugates for gold nanoparticles tro2020), under the PORTUGAL 2020 Partnership Agreement, through anticancer drugs: where the seed meets the soil? Biomedicines 9 (2021) 74, the European Regional Development Fund. In addition, support through https://doi.org/10.3390/biomedicines9010074. the project IF/00894/2015 and within the scope of the project CICECO- [15] A.H. Soloway, W. Tjarks, B.A. Barnum, F.-G. Rong, R.F. Barth, I.M. Codogni, J. G. Wilson, The chemistry of neutron capture therapy, Chem. Rev. 98 (1998) Aveiro Institute of Materials, UIDB/50011/2020, UIDP/50011/2020 & 1515–1562, https://doi.org/10.1021/cr941195u. LA/P/0006/2020, financed by national funds through the FCT/MEC [16] P.A. Zahl, F.S. Cooper, J.R. Dunning, Some in vivo effects of localized nuclear (PIDDAC) is gratefully acknowledged. We acknowledge funding by disintegration products on a transplantable mouse sarcoma, Proc. Natl. Acad. Sci. INFN (CSN5)-project ENTER_BNCT. ICMAB and ICN2 acknowledge United States Am. 26 (1940) 589–598, https://doi.org/10.1073/pnas.26.10.589. [17] P.A. Zahl, F.S. Cooper, Physical and biological considerations in the use of slow financial support from the Spanish Ministry of Economy and Competi­ neutrons for cancer therapy, Radiology 37 (1941) 673–682, https://doi.org/ tiveness (Spain), through the “Severo Ochoa” Programme for Centres of 10.1148/37.6.673. Excellence in R&D (CEX2019-000917-S and CEX2021-001214-S [18] P.A. Zahl, F.S. Cooper, Localization of lithium in tumor tissue as a basis for slow neutron therapy, Science 93 (1941) 64–65, https://doi.org/10.1126/ respectively). ICN2 is supported by CERCA programme. We acknowl­ science.93.2403.64. edge funding from Generalitat de Catalunya (2021-SGR-00439, 2017 [19] R. Núñez, M. Tarrés, A. Ferrer-Ugalde, F.F. de Biani, F. Teixidor, Electrochemistry SGR 327). M.Ll. has carried out this work in the framework of the and photoluminescence of icosahedral carboranes, boranes, metallacarboranes, and their derivatives, Chem. Rev. 116 (2016) 14307–14378, https://doi.org/ Doctoral Degree Program in Materials Science of the Universitat 10.1021/acs.chemrev.6b00198. Autònoma de Barcelona. We acknowledge fruitful discussions with [20] D. Olid, R. Núñez, C. Viñas, F. Teixidor, Methods to produce B–C, B–P, B–N and B–S Manuel Altabas and support with the XPS analysis by Guillaume bonds in boron clusters, Chem. Soc. Rev. 42 (2013) 3318–3336, https://doi.org/ 10.1039/C2CS35441A. Sauthier. [21] L.M. Rendina, Can lithium salts herald a new era for neutron capture therapy? J. Med. Chem. 53 (2010) 8224–8227, https://doi.org/10.1021/jm100836m. Appendix A. Supplementary data [22] L. Tondo, M. Alda, M. Bauer, V. Bergink, P. Grof, T. Hajek, U. Lewitka, R.W. Licht, M. Manchia, B. Müller-Oerlinghausen, R.E. Nielsen, M. Selo, C. Simhandl, R. J. Baldessarini, For the I.G. for S. of L. (IGSLi), Clinical use of lithium salts: guide Supplementary data to this article can be found online at https://doi. for users and prescribers, Int. J. Bipolar Disord. 7 (2019) 16, https://doi.org/ org/10.1016/j.carbon.2023.03.034. 10.1186/s40345-019-0151-2. [23] D. Draaisma, Lithium: the gripping history of a psychiatric success story, Nature 572 (2019) 584–586. References [24] D. Li, Y. Zhang, J. Xu, F. Yoshino, H. Xu, X. Chen, L. Zhao, Surface-engineered carbon nanohorns as a theranostic nanodevice for photoacoustic imaging and effective radiochemotherapy of cancer, Carbon N. Y. 180 (2021) 185–196, https:// [1] A. Curt, H.J.A. Van Hedel, D. Klaus, V. Dietz, Recovery from a spinal cord injury: doi.org/10.1016/j.carbon.2021.04.073. significance of compensation, neural plasticity, and repair, J. Neurotrauma 25 [25] Y. Maeda, E. Hirata, Y. Takano, N. Sakaguchi, N. Ushijima, A. Saeki, S. Kimura, (2008) 677–685, https://doi.org/10.1089/neu.2007.0468. K. Shibata, M. Yudasaka, A. Yokoyama, Stable aqueous dispersions of carbon [2] M. Suzuki, Boron neutron capture therapy (BNCT): a unique role in radiotherapy nanohorns loaded with minocycline and exhibiting antibacterial activity, Carbon with a view to entering the accelerator-based BNCT era, Int. J. Clin. Oncol. 25 N. Y. 166 (2020) 36–45, https://doi.org/10.1016/j.carbon.2020.04.040. (2020) 43–50, https://doi.org/10.1007/s10147-019-01480-4. [26] S. Guo, Z. Song, D.-K. Ji, G. Reina, J.-D. Fauny, Y. Nishina, C. Ménard-Moyon, [3] K. Hirose, A. Konno, S. Yoshimoto, K. Ono, N. Otsuki, J. Hatazawa, J. Hiratsuka, A. Bianco, Combined photothermal and photodynamic therapy for cancer Y. Takai, A prospective phase 2 trial of accelerator-based boron neutron capture treatment using a multifunctional graphene oxide, Pharmaceutics 14 (2022), therapy for recurrent and locally advanced head and neck cancer: initial report of https://doi.org/10.3390/pharmaceutics14071365. treatment outcome, Int. J. Radiat. Oncol. 105 (2019) E374, https://doi.org/ [27] V. Meunier, C. Ania, A. Bianco, Y. Chen, G.B. Choi, Y.A. Kim, N. Koratkar, C. Liu, J. 10.1016/j.ijrobp.2019.06.1625. M.D. Tascon, M. Terrones, Carbon science perspective in 2022: current research [4] R.F. Barth, Z. Zhang, T. Liu, A realistic appraisal of boron neutron capture therapy and future challenges, Carbon N. Y. 195 (2022) 272–291, https://doi.org/ as a cancer treatment modality, Cancer Commun. 38 (2018) 36, https://doi.org/ 10.1016/j.carbon.2022.04.015. 10.1186/s40880-018-0280-5. [28] S.I. Amaral, R. Costa-Almeida, I.C. Gonçalves, F.D. Magalhães, A.M. Pinto, Carbon [5] T.-Z. Yuan, S.-Q. Xie, C.-N. Qian, Boron neutron capture therapy of cancer: critical nanomaterials for phototherapy of cancer and microbial infections, Carbon N. Y. issues and future prospects, Thorac. Cancer. 10 (2019) 2195–2199, https://doi. 190 (2022) 194–244, https://doi.org/10.1016/j.carbon.2021.12.084. org/10.1111/1759-7714.13232. [29] C.J. Serpell, K. Kostarelos, B.G. Davis, Can carbon nanotubes deliver on their [6] N.V. Gubanova, A.R. Tsygankova, E.L. Zavjalov, A.V. Romashchenko, Y.L. Orlov, promise in biology? Harnessing unique properties for unparalleled applications, Biodistribution of 10B in glioma orthotopic xenograft mouse model after injection ACS Cent. Sci. 2 (2016) 190–200, https://doi.org/10.1021/acscentsci.6b00005. of L-para-boronophenylalanine and sodium borocaptate, Biomedicines 9 (2021) [30] S.Y. Hong, G. Tobias, K.T. Al-Jamal, B. Ballesteros, H. Ali-Boucetta, S. Lozano- 722, https://doi.org/10.3390/biomedicines9070722. Perez, P.D. Nellist, R.B. Sim, C. Finucane, S.J. Mather, M.L.H. Green, K. Kostarelos, [7] R.F. Barth, P. Mi, W. Yang, Boron delivery agents for neutron capture therapy of B.G. Davis, Filled and glycosylated carbon nanotubes for in vivo radioemitter cancer, Cancer Commun. 38 (2018) 35, https://doi.org/10.1186/s40880-018- localization and imaging, Nat. Mater. 9 (2010) 485–490, https://doi.org/10.1038/ 0299-7. London, England. nmat2766. [8] C. Viñas, R. Núñez, I. Bennour, F. Teixidor, Periphery decorated and core initiated [31] H. Ge, P.J. Riss, V. Mirabello, D.G. Calatayud, S.E. Flower, R.L. Arrowsmith, T. neutral and polyanionic borane large molecules: forthcoming and promising D. Fryer, Y. Hong, S. Sawiak, R.M.J. Jacobs, S.W. Botchway, R.M. Tyrrell, T. properties for medicinal applications, Curr. Med. Chem. 26 (2019) 5036–5076, D. James, J.S. Fossey, J.R. Dilworth, F.I. Aigbirhio, S.I. Pascu, Behavior of https://doi.org/10.2174/0929867326666190603123838. supramolecular assemblies of radiometal-filled and fluorescent carbon [9] M. Nuez-Martinez, C.I.G. Pinto, J.F. Guerreiro, F. Mendes, F. Marques, A. Muñoz- nanocapsules in vitro and in vivo, Chem 3 (2017) 437–460, https://doi.org/ Juan, J.A. Xavier, A. Laromaine, V. Bitonto, N. Protti, S.G. Crich, F. Teixidor, 10.1016/j.chempr.2017.06.013. C. Viñas, Cobaltabis(dicarbollide) ([o-COSAN]− ) as multifunctional [32] C.J. Serpell, R.N. Rutte, K. Geraki, E. Pach, M. Martincic, M. Kierkowicz, S. De chemotherapeutics: a prospective application in boron neutron capture therapy Munari, K. Wals, R. Raj, B. Ballesteros, G. Tobias, D.C. Anthony, B.G. Davis, Carbon (BNCT) for glioblastoma,, Cancers 13 (2021), https://doi.org/10.3390/ nanotubes allow capture of krypton, barium and lead for multichannel biological cancers13246367. X-ray fluorescence imaging, Nat. Commun. 7 (2016), https://doi.org/10.1038/ [10] S.N. Bhatia, X. Chen, M.A. Dobrovolskaia, T. Lammers, Cancer nanomedicine, Nat. ncomms13118. Rev. Cancer 22 (2022) 550–556, https://doi.org/10.1038/s41568-022-00496-9. 158 G. Gonçalves et al. Carbon 208 (2023) 148–159 [33] J.T.-W. Wang, R. Klippstein, M. Martincic, E. Pach, R. Feldman, M. Šefl, Y. Michel, S. Thygesen, T. Vegge, L. Vilhelmsen, M. Walter, Z. Zeng, K.W. Jacobsen, The D. Asker, J.K. Sosabowski, M. Kalbac, T. Da Ros, C. Ménard-Moyon, A. Bianco, atomic simulation environment—a Python library for working with atoms, J. Phys. I. Kyriakou, D. Emfietzoglou, J.-C. Saccavini, B. Ballesteros, K.T. Al-Jamal, Condens. Matter 29 (2017), 273002, https://doi.org/10.1088/1361-648x/aa680e. G. Tobias, Neutron activated 153Sm sealed in carbon nanocapsules for in vivo [46] L. Bellarosa, E. Bakalis, M. Melle-Franco, F. Zerbetto, Interactions in concentric imaging and tumor radiotherapy, ACS Nano 14 (2020), https://doi.org/10.1021/ carbon nanotubes: the radius vs the chirality angle contributions, Nano Lett. 6 acsnano.9b04898. (2006) 1950–1954, https://doi.org/10.1021/nl061066g. [34] J.T.-W. Wang, C. Spinato, R. Klippstein, P.M. Costa, M. Martincic, E. Pach, A. [47] S. Bortolussi, I. Postuma, N. Protti, L. Provenzano, C. Ferrari, L. Cansolino, P. Ruiz de Garibay, C. Ménard-Moyon, R. Feldman, Y. Michel, M. Šefl, I. Kyriakou, P. Dionigi, O. Galasso, G. Gasparini, S. Altieri, S.-I. Miyatake, S.J. González, D. Emfietzoglou, J.-C. Saccavini, B. Ballesteros, G. Tobias, A. Bianco, K.T. Al-Jamal, Understanding the potentiality of accelerator based-boron neutron capture therapy Neutron-irradiated antibody-functionalised carbon nanocapsules for targeted for osteosarcoma: dosimetry assessment based on the reported clinical experience, cancer radiotherapy, Carbon N. Y. 162 (2020) 410–422, https://doi.org/10.1016/ Radiat. Oncol. 12 (2017) 130, https://doi.org/10.1186/s13014-017-0860-6. j.carbon.2020.02.060. [48] I. Postuma, S. Bortolussi, N. Protti, F. Ballarini, P. Bruschi, L. Ciani, S. Ristori, [35] A. Gajewska, J.T.-W. Wang, R. Klippstein, M. Martincic, E. Pach, R. Feldman, J.- L. Panza, C. Ferrari, L. Cansolino, S. Altieri, An improved neutron autoradiography C. Saccavini, G. Tobias, B. Ballesteros, K.T. Al-Jamal, T. Da Ros, Functionalization set-up for (10)B concentration measurements in biological samples, Reports Pract. of filled radioactive multi-walled carbon nanocapsules by arylation reaction for in Oncol. Radiother. J. Gt. Cancer Cent. Pozn. Polish Soc. Radiat. Oncol. 21 (2016) vivo delivery of radio-therapy, J. Mater. Chem. B. 10 (2022) 47–56, https://doi. 123–128, https://doi.org/10.1016/j.rpor.2015.10.006. org/10.1039/D1TB02195H. [49] S. Altieri, S. Bortolussi, P. Bruschi, P. Chiari, F. Fossati, S. Stella, U. Prati, [36] M. Kierkowicz, E. Pach, A. Santidrián, S. Sandoval, G. Gonçalves, E. Tobías-Rossell, L. Roveda, A. Zonta, C. Zonta, C. Ferrari, A. Clerici, R. Nano, T. Pinelli, Neutron M. Kalbáč, B. Ballesteros, G. Tobias, Comparative study of shortening and cutting autoradiography imaging of selective boron uptake in human metastatic tumours, strategies of single-walled and multi-walled carbon nanotubes assessed by scanning Appl. Radiat. Isot. Incl. Data, Instrum. Methods Use Agric. Ind. Med. 66 (2008) electron microscopy, Carbon N. Y. 139 (2018) 922–932, https://doi.org/10.1016/ 1850–1855, https://doi.org/10.1016/j.apradiso.2008.05.007. j.carbon.2018.06.021. [50] K. Kostarelos, The long and short of carbon nanotube toxicity, Nat. Biotechnol. 26 [37] S. Utsumi, J. Miyawaki, H. Tanaka, Y. Hattori, T. Itoi, N. Ichikuni, H. Kanoh, (2008) 774–776, https://doi.org/10.1038/nbt0708-774. M. Yudasaka, S. Iijima, K. Kaneko, Opening mechanism of internal nanoporosity of [51] A. Kovtun, E. Treossi, N. Mirotta, A. Scidà, A. Liscio, M. Christian, F. Valorosi, single-wall carbon nanohorn, J. Phys. Chem. B 109 (2005) 14319–14324, https:// A. Boschi, R.J. Young, C. Galiotis, I.A. Kinloch, V. Morandi, V. Palermo, doi.org/10.1021/jp0512661. Benchmarking of graphene-based materials: real commercial products versus ideal [38] G. Brown, S.R. Bailey, M. Novotny, R. Carter, E. Flahaut, K.S. Coleman, J. graphene, 2D Mater. 6 (2019), 25006, https://doi.org/10.1088/2053-1583/ L. Hutchison, M.L.H. Green, J. Sloan, High yield incorporation and washing aafc6e. properties of halides incorporated into single walled carbon nanotubes, Appl. Phys. [52] M. Kierkowicz, J.M. González-Domínguez, E. Pach, S. Sandoval, B. Ballesteros, A 76 (2003) 457–462, https://doi.org/10.1007/s00339-002-2040-1. T. Da Ros, G. Tobias, Filling single-walled carbon nanotubes with lutetium [39] M. Martincic, S. Vranic, E. Pach, S. Sandoval, B. Ballesteros, K. Kostarelos, chloride: a sustainable production of nanocapsules free of nonencapsulated G. Tobias, Non-cytotoxic carbon nanocapsules synthesized via one-pot filling and material, ACS Sustain. Chem. Eng. 5 (2017) 2501–2508, https://doi.org/10.1021/ end-closing of multi-walled carbon nanotubes, Carbon N. Y. 141 (2019) 782–793, acssuschemeng.6b02850. https://doi.org/10.1016/j.carbon.2018.10.006. [53] A.V. Crewe, J. Wall, J. Langmore, Visibility of single atoms, Science 168 (1970) [40] J.M. González-Domínguez, A. Santidrián, A. Criado, C. Hadad, M. Kalbáč, T. Da 1338–1340, https://doi.org/10.1126/science.168.3937.1338. Ros, Multipurpose nature of rapid covalent functionalization on carbon nanotubes, [54] P.D. Nellist, S.J. Pennycook, Science 274 (1996) 413. Chem. Eur J. 21 (2015) 18631–18641, https://doi.org/10.1002/chem.201503085. [55] R.J. Kashtiban, C.E. Patrick, Q. Ramasse, R.I. Walton, J. Sloan, Picoperovskites, [41] S. Grimme, C. Bannwarth, P. Shushkov, A robust and accurate tight-binding The smallest conceivable isolated halide perovskite structures formed within quantum chemical method for structures, vibrational frequencies, and noncovalent carbon nanotubes, Adv. Mater. (2022), 2208575, https://doi.org/10.1002/ interactions of large molecular systems parametrized for all spd-block elements (Z adma.202208575. = 1–86), J. Chem. Theor. Comput. 13 (2017) 1989–2009, https://doi.org/ [56] S. Stonemeyer, M. Dogan, J.D. Cain, A. Azizi, D.C. Popple, A. Culp, C. Song, 10.1021/acs.jctc.7b00118. P. Ercius, M.L. Cohen, A. Zettl, Targeting one- and two-dimensional Ta–Te [42] S. Piovesana, D. Iglesias, M. Melle-Franco, S. Kralj, C. Cavaliere, M. Melchionna, structures via nanotube encapsulation, Nano Lett. 22 (2022) 2285–2292, https:// A. Laganà, A.L. Capriotti, S. Marchesan, Carbon nanostructure morphology doi.org/10.1021/acs.nanolett.1c04615. templates nanocomposites for phosphoproteomics, Nano Res. 13 (2020) 380–388, [57] L. Shao, G. Tobias, Y. Huh, M.L.H. Green, Reversible filling of single walled carbon https://doi.org/10.1007/s12274-020-2620-4. nanotubes opened by alkali hydroxides, Carbon N, Y (Dayt. Ohio) 44 (2006) [43] B. Hourahine, B. Aradi, V. Blum, F. Bonafé, A. Buccheri, C. Camacho, C. Cevallos, 2855–2858, https://doi.org/10.1016/j.carbon.2006.06.010. M.Y. Deshaye, T. Dumitrică, A. Dominguez, S. Ehlert, M. Elstner, T. van der Heide, [58] M. Kierkowicz, E. Pach, J. Fraile, C. Domingo, B. Ballesteros, G. Tobias, The role of J. Hermann, S. Irle, J.J. Kranz, C. Köhler, T. Kowalczyk, T. Kubař, I.S. Lee, temperature on the degree of end-closing and filling of single-walled carbon V. Lutsker, R.J. Maurer, S.K. Min, I. Mitchell, C. Negre, T.A. Niehaus, A.M. nanotubes, Nanomaterials 11 (2021) 3365, https://doi.org/10.3390/ N. Niklasson, A.J. Page, A. Pecchia, G. Penazzi, M.P. Persson, J. Řezáč, C. nano11123365. G. Sánchez, M. Sternberg, M. Stöhr, F. Stuckenberg, A. Tkatchenko, V.W. -z. Yu, [59] M.V.A.U.-K. Kharlamova, ChristianTI - applications of filled single-walled carbon T. Frauenheim, DFTB+, a software package for efficient approximate density nanotubes: progress, challenges, and perspectives, No title, Nanomaterials 11 functional theory based atomistic simulations, J. Chem. Phys. 152 (2020), 124101, (2021) 2863, https://doi.org/10.3390/nano11112863. https://doi.org/10.1063/1.5143190. [60] B. He, Y. Shi, Y. Liang, A. Yang, Z. Fan, L. Yuan, X. Zou, X. Chang, H. Zhang, [44] D. Stradi, L. Jelver, S. Smidstrup, K. Stokbro, Method for determining optimal X. Wang, W. Dai, Y. Wang, Q. Zhang, Single-walled carbon-nanohorns improve supercell representation of interfaces, J. Phys. Condens. Matter 29 (2017), 185901, biocompatibility over nanotubes by triggering less protein-initiated pyroptosis and https://doi.org/10.1088/1361-648x/aa66f3. apoptosis in macrophages, Nat. Commun. 9 (2018) 2393, https://doi.org/ [45] A. Hjorth Larsen, J. Jørgen Mortensen, J. Blomqvist, I.E. Castelli, R. Christensen, 10.1038/s41467-018-04700-z. M. Dułak, J. Friis, M.N. Groves, B. Hammer, C. Hargus, E.D. Hermes, P.C. Jennings, [61] M. Florensa, M. Llenas, E. Medina-Gutiérrez, S. Sandoval, G. Tobías-Rossell, Key P. Bjerre Jensen, J. Kermode, J.R. Kitchin, E. Leonhard Kolsbjerg, J. Kubal, parameters for the rational design, synthesis, and functionalization of K. Kaasbjerg, S. Lysgaard, J. Bergmann Maronsson, T. Maxson, T. Olsen, biocompatible mesoporous silica nanoparticles, Pharmaceutics 14 (2022), https:// L. Pastewka, A. Peterson, C. Rostgaard, J. Schiøtz, O. Schütt, M. Strange, K. doi.org/10.3390/pharmaceutics14122703. 159