Introduction

Graphene aerogels (GAs) have attracted significant attention recently since they are ultra-lightweight, mechanically robust, electrically conductive, perform well under harsh environment and as a result have found good use in various applications1. The GAs can also be used as reinforcing agents in composites, in sensors and nano-electronics, energy storage2, as cell growth promoters3, absorbers for the removal of food toxins4, in catalysis and biomedical applications among others1. Other materials such as hexagonal boron nitride (hBN) have also attracted significant attention for the fabrication of aerogels5 since their distinct properties as compared to GAs, pave the way for the development of new applications. For example, boron nitride based aerogels (BNAs) have been extensively used for the removal of organic and inorganic pollutants from aqueous solutions with high efficiency6. Moreover, BNAs possess high thermal conductivity and they remain stable up to 1000o C which make them attractive materials for applications required to perform reliably under harsh environments7. Besides graphene and boron nitride based aerogels, the area is highly active and other materials have also been tested, including silica8,9 and SnO210 based aerogels that show attractive properties and have been used in various applications.

In an attempt to combine and exploit the distinct properties of dissimilar 2D materials, the fabrication of 2D hybrids could, under certain circumstances, impart new multi-functionalities and therefore has recently emerged as a very intense research field11,12,13,14,15,16,17. In this regard, hybrid aerogels (HAs) have started to gain attention and some efforts have already been reported in the literature for aerogels composed of combined graphene/boron nitride nano-sheets11,12, graphene/ boron nitride nanotubes13, graphene/MoS214,15,16 and others17. In particular, BNAs exhibit high thermal conductivity but low electrical conductivity,18 therefore, they are only suitable for applications as insulators. In contrast, reduced graphene oxide has both high thermal and moderate electrical conductivity and therefore its field of application differs than hBN. Both materials possess attractive mechanical properties, they are highly compressible and perform well as absorbers of gas and oil pollutants. It would be of great interest to create hybrid aerogels that retain the properties of the standalone materials or even develop hybrids that could outperform the efficiency of the homostructured materials.

One particular application of the aerogels is their use as volatile organic compounds (VOC) absorbers for the protection of museum artefacts such as paintings and other objects. It is worth noting that most artefacts owned by museums are often stored under hostile environmental conditions in deposits (e.g. in crates and boxes). It has been estimated that the largest museums typically display only about 5% of their collection at any time. This roughly 5% is usually rotated among the most important works the museums have, while other less significant works may never leave the storage area. Additionally, some artworks may be displayed but need preservation frequently in order that they survive, like works based on paper, which will fade if exposed to light. Both carbon and hBN aerogels have been tested as oil and organic absorbers19. Bi et al.20 examined the efficiency of spongy graphene for oil absorption by floating the sample in liquid pollutants such as commercial petroleum products, fats and ketones, and the results showed high speed of absorption and efficient recovery. Similar experiments have been performed for thermally cross-linked Poly(acrylic acid)/rGO of low porosity as absorbers of various oils with high efficiency21, as well for pure graphene aerogels with density less than ~ 9 mg/cm322. An extensive discussion on the subject regarding carbon based absorbers can be found in reference19 and for hBN based materials in reference6. Boron nitride based porous materials have also been tested for the removal of substances from water. A foam like hBN structure was developed with density of ~ 30.4 mg/cm3 and proved very efficient for the removal/separation of a number of oils and chemicals by immersing the porous materials into contaminated water23. Similar experiments performed for another sponge like material made by few-layer hBN-coated melamine sponge which showed to possess very good recyclability performance24.

Herein, we fabricate hybrid aerogels at various mix ratios composed of reduced graphene oxide (rGO) and hexagonal boron nitride (hBN) platelets using a freeze-drying approach. The method is cost effective and can be easily used for the fabrication of large-scale quantities, while it can be extended to the fabrication of other 2D materials as well. In this work we examine in detail the mechanical properties of the HAs as well the changes in the electrical conductivity induced by the presence of hBN. Furthermore, we investigate the efficiency of the proposed HAs as gas absorbers of formaldehyde, hydrochloric acid and acetic acid, as well their capacity for water vapour absorption. The results presented below show that the rGO/hBN hybrid aerogels possess good mechanical properties and reasonable electrical conductivities but, most importantly, certain combinations of these hybrids exhibit extremely high gas absorption capabilities.

Experimental

The approach presented previously by Hong et al.25 was followed herein for the fabrication of the rGO aerogels. Aqueous solution of GO was prepared by modified Hummer’s method26,27 and was subsequently diluted in water to obtain a concentration of 1 mg/mL. Hypophosphorous acid (H3PO2) and iodine (I2) of weight ratio GO: H3PO2:I2 equal to 1:100:10, were then added as the chemically reducing agents28. Subsequently the solution was placed in a furnace and heated to 80 °C for 8 h, resulting in a uniform gelation of the GO. The sample was then rinsed with water until a pH equal to 5, followed by freeze-drying for 48 h. For the fabrication of the HAs, another solution with boron nitride platelets of 1 mg/mL concentration was prepared in ethanol–water (50–50% w/w) and stirred for 10 min. Solutions with mixed GO and hBN with various concentrations of 0.9/0.1, 0.7/0.3 and 0.5/0.5 mg/mL were stirred and then the same steps as for the case of the pure GO were followed. We note that the addition of ethanol significantly stabilizes the mixed solution29 and assists to achieve uniform dispersions as seen in Fig. 1b. When we used only water, we observed that the hBN material tends to separate from the mixed solution and even though gelation still occurs, there is no uniform distribution of the hBN platelets in the formed aerogel. This bottleneck is attributed to the absence of polar groups to the surface of hBN13. It is noted that this approach cannot be used for the preparation of pure hBN aerogels, which points out to the crucial role of the presence of the GO which stabilizes the mixture. This is due to the gelation mechanism that involves the simultaneous reduction and self-assembly of the GO sheets, based on chemical reactions between the oxygen functional groups present at the surface of the GO and the chemical agents28. Thus, the hBN platelets that do not contain such groups on their surface simply attach to the GO sheets. The only complication is that the gelation time might be increased in the mixed solutions as discussed previously13. The density of the aerogels is found to be in the range of 13.9–23.4 mg/cm−3 without significant variation between samples.

Figure 1
figure 1

(a) An ultra-light weight rGO aerogel which is shown to be supported by just the stamen of a flower inside the lab, (b) starting solutions for the development of the HAs: from left to right, neat rGO, rGO-hBN 90/10, rGO-hBN 70/30 and rGO-hBN 50/50 and (c) the as-prepared hybrid aerogels with different amounts of graphene oxide and hBN with the same order as in (b).

The surface analysis measurements were performed in a UHV chamber (P ~ 5 × 10 − 10 mbar) equipped with a SPECS Phoibos 100-1D-DLD hemispherical electron analyser and a non-monochromatized dual-anode Mg/Al x-ray source for XPS. The XP Spectra were recorded with MgKa at 1253.6 eV photon energy and an analyser pass energy of 10 eV giving a Full Width at Half Maximum (FWHM) of 0.85 eV for Ag3d5/2 line. The analysed area was a spot with 3 mm diameter. The atomic ratios were calculated from the intensity (peak area) of the XPS peaks weighted with the corresponding relative sensitivity factors (RSF) derived from the Scofield cross-section taking into account the electron transport properties of the matrix, [namely the inelastic mean free path (IMFP) λi and the elastic-scattering correction factor Q, depend mainly on the corresponding electron kinetic energy (KE)] and the energy analyser transmission function. For spectra collection and treatment, including fitting, the commercial software SpecsLab Prodigy (by Specs GmbH, Berlin) was used. The XPS peaks were deconvoluted with sum Gaussian-Lorenzian peaks after a Shirtey type background subtraction.

The gas absorption test was conducted in static conditions into a closed glass desiccator with an excess of pollutant, using a saturated vapour stream at room temperature. 100 mL of either formaldehyde (37 wt% in H2O), hydrochloric acid or acetic acid were used as gas source each time. For the humidity measurements, vials full of water were placed similarly to the gases, while the relative humidity (RH) was measured independently and was found to be ~ 97%. All samples were initially dried at 200 °C for 2 h in order to remove their absorbed humidity and weighed into a high accuracy weight meter to measure their dried mass. Afterwards, the aerogels were loaded in a glass petri dish which was mounted inside the desiccator. The desiccator with the aerogels and the fuming gas was constantly inside a fume hood. Periodically, the mass of each sample was recorded by the weight meter, exactly next to the fume hood minimizing the exposure of the aerogels to the environment. The regeneration of the graphene and the graphene-hBN aerogels was performed by using a common electric hair dryer. During the regeneration process, the formed volatile vapour was removed with the assistance of the hood, as reported by Li et al.30.

Results and discussion

The approach presented previously by Hong et al.25 was followed herein for the fabrication of graphene based rGO aerogels and is described in detail in the experimental section. For the fabrication of the HAs, another solution with boron nitride platelets was prepared in ethanol–water and mixed with the GO in various concentrations of 0.9/0.1, 0.7/0.3 and 0.5/0.5 mg/mL, and the same procedure as in the case of the GA was pursued. We note that the addition of ethanol significantly stabilizes the mixed solution29 and assists to achieve uniform dispersions as seen in Fig. 1b. The density of the aerogels is found to be in the range of 13.9–23.4 mg/cm−3 without significant variation between samples. The advantages of the present approach for the fabrication of hybrid aerogels are that it can be easily used for scalable fabrication since it is less energy demanding than other approaches31. It must be stressed that this synthesis method cannot be used for the fabrication of pure hBN aerogels.

In Fig. 1, the as-prepared hybrid aerogels are shown and the corresponding solutions prior to gelation and subsequent freeze-drying. A variety of techniques was employed for the characterization of the HAs including X-ray photoelectron spectroscopy (XPS), X-ray diffraction (XRD), Raman spectroscopy and scanning electron microscopy (SEM). The Survey XPS Scans (Fig. 2a) show the presence of C, O and P atoms in both pure GA and HA samples, while the trace of N and B is present on the HAs sample surface. The deconvoluted XPS C1s is shown in Fig. 2b,c. By analysing the peak (Table 1) we detect the presence of sp2 and sp2 hydridization and also of oxides in the form of epoxides, hydroxides, carbonyls and carboxyls due to pi–pi* transition loss peak32. From the peak areas of the C1s, the percentage (%) component concentration is calculated and from the peak intensities of C1s, O1s, P2p and N1s the (%) relative atomic ratio is obtained. Both sets of results are shown in Table 1. In order to calculate the C:O atomic ratio in each specimen, the oxygen concentration due to P2O5 chemical state is subtracted. The results are for the rGO-hBN (50/50) C:O = 5.6, and C:O = 10 for the pure GA, suggesting the effective reduction of the GO to rGO33. The C:O ratio for the 50/50 is almost the half of the pure GA which shows that the presence of hBN affects the reduction of the GO.

Figure 2
figure 2

(a) Survey Scans of the hBN and HPO samples. (b,c) Deconvoluted C1s XPS peaks of the hBN and HPO samples, respectively.

Table 1 Percentage of C1s component concentration derived from the C1s peak deconvolution (Fig. 2) and relative atomic ratio C:O:P of the GA and C:O:P:N:B of the HA sample.

In Fig. 3 Raman and XRD measurements are shown for the rGO and HA samples. Both the Raman and XRD measurements prove the presence of both materials in the hybrid aerogels. The XRD plots show a broad peak at 2θ = 26° which corresponds to the (002) plane of graphite structure, as reported elsewhere34. The HAs show graphite peak at 2θ = 26° and also a clear peak at 27.3°, which are attributed to the (002) diffraction of hBN12. The position of the Raman peaks for the GO are ~ 1590 cm−1 and 1356 cm−1 corresponding to the G and D phonons, respectively35. The presence of hBN is also evident in the HA from an intense peak at 1366 cm−1 originating from the in-plane E2g phonon mode36. The intensity of the peak of the hBN is large due to the relatively large thickness (~ 2–3 μm) of the platelets.

Figure 3
figure 3

(a,c) XRD and Raman spectra of the rGO and the corresponding spectra (b,d) for the rGO-hBN aerogels samples. The as-made rGO aerogel exhibits a weak broad peak, characteristic of an amorphous material, at 2θ = 26°, which corresponds to the (002) plane of graphite structure, as reported elsewhere34. The hybrid rGO–hBN aerogel shows mainly the before-mentioned rGO peak at 2θ = 26° and also a clear peak at 27.3° which is attributed to the (002) diffraction peak of hBN12. The characteristic G and D peaks are clearly seen in the Raman spectra for the rGO and also the presence of hBN is confirmed by the peak of the E2g phonon at ~ 1366 cm−1.

SEM images for all samples with various mixed ratios are given in Fig. 4. The hBN platelets are clearly distinguished and seem to be either attached to the surface of the rGO layers or wrapped around them, creating a robust macro-scale porous network. We also observe from the SEM images that the magnitude of the pores tends to increase with the increase in the amount of hBN in the HA samples. This can be explained by considering the amount of the rGO in each structure and the relative volume of the resulted aerogel. The 3D porous assembled structure depends on the available amount of rGO. The resulted aerogels show similar macroscale volumes in all cases. Thus, with decreasing rGO, less amount of material assembles and occupies the same volume (since the hBN is wrapped around the rGO and does not contribute to the formation of the 3D network), as seen in the optical pictures presented in Fig. 1. Moreover, this is also manifested in the average density values given in Table 2, which is less than 30 mg/cm−3 for all cases. In fact, this value is a prerequisite for classifying the 3D porous structure as an aerogel12. The structural characteristics of the integration of the rGO with the hBN, such as the attachment and wrapping of the hBN to the rGO, agrees very well with the results reported in other studies11,12.

Table 2 The average density of the samples for all cases. In the second column the density of specific samples for which the electrical conductivity was measured and is given the last column.

The mechanical properties under compression for the hybrid structures of variable rGO/ hBN content were examined in detail and compared to those of the pure rGO aerogel. In Fig. 5 representative stress–strain curves for all tested samples are presented and in Fig. 5b the modulus along with the electrical conductivity of all cases. All samples compressed up to a maximum compressive strain of ~ 80–90% without failure, followed by recording the unloading behaviour. The stress–strain response consists of three discrete stages, characteristic of the compressive behaviour of aerogels. Initially a linear Hookean behaviour, which holds roughly up to 10%, is observed corresponding to the elastic bending of the cell walls. For all cases the linear stiffness is similar with values in the range of ~ 64–80 kPa. The initial linear behaviour from which the Young’s modulus is extracted, is followed by a non-linear regime of much lower modulus. During this regime, densification of the compressed walls of the aerogel sets off, until a compressive strain of 60−70%. For higher compression, the stress increases rapidly due to the resulting densification and closing of the pores of the 3D structure. Some differences in the compressive behaviour are plausibly expected due to the non-perfect cylindrical geometry of the samples and the variation in the density, but overall, the results are consistent within the experimental error. The differences in the energy dissipation are also attributed to the variation in the density of the samples. We note that the similar mechanical properties (maximum compressive stresses and modulus) of the present HAs with the pure GA reflect the efficiency of the fabrication procedure and the effective synergy of the two materials. It seems therefore that using rGO to support mechanically the hBN is a very effective strategy also for fabrication purposes, since high aspect ratio HAs can be easily produced and more importantly the distinct properties of both materials can be preserved as discussed below.

In a previous study, similar HAs with a smaller density of ~ 3.6 mg/cm−3 were tested and a compressive stress of ~ 1.5 kPa was measured at 50% of strain11. Elsewhere, rGO aerogels reinforced with boron nitride nanotubes of total density ~ 16 mg/cm−3, reached stress of ~ 12 kPa at ~ 75% of compressive strain13. We note that the mechanical properties of aerogels are linearly related to the density of the samples37,38, and the modulus for the GA of the present study agrees well with the analysis of reference37. Thus, the present HAs possess better mechanical performance as compared to reported results for other heterostructures11,13 since they have similar density or slightly higher, and in our case they reach much higher compressive stresses without failure.

In Fig. 4b, the electrical conductivity for all cases is presented. In Table 1 the electrical conductivity and the density of the corresponding specimens measured are given. The neat rGO aerogel presents a relatively high conductivity with value of ~ 40.4 S/m, indicating the effective reduction of the GO in agreement with the results from the XPS measurements. This value is close to the highest reported conductivities in the range of 50–87 S/m for graphene aerogels with similar densities37,39. The 90/10 hybrid exhibits almost half of this value of ~ 23.3 S/m, and with further increment of the amount of the hBN the conductivities do not vary a lot within the experimental error (Table 1). Overall, any differences are attributed to density variation, the extend of the reduction of the rGO, as well as, of course the significant differences in electrical conductivity between the two constituents. Effectively, this means that the conductivity of the pure GA and the 50/50 sample should differ by a factor of 2, assuming they possess the same density. Instead, from Table 1 we observe that despite these two samples have similar density, the conductivity of 50/50 is smaller by a factor of 2.6. This additional decrease is due to the less effective reduction of the GO to rGO with the presence of hBN as evident by the XPS measurements. At any rate, the electrical conductivity is considered reasonably high for all cases. The electrical conductivity along with the excellent thermal behaviour and stability of hBN at high temperatures12,40, can be exploited in a variety of applications for these aerogels, for example as electro-thermal harvesters41,42. The results are summarized in Fig. 5b.

Figure 4
figure 4

SEM images of the (a) rGO aerogel, (b) 90/10 HA, (c) 70/30 HA and (d) 50/50 HA. In (e) a zoom of the rGO aerogel is presented and in (f) a zoon of hybrid aerogel which clearly show the presence of hBN platelets in the HA samples and have size of a few microns in length and about ~ 3–5 microns thickness. The scale bar is 20 μm for (ad) while for (e,f) is 4 μm.

Figure 5
figure 5

(a) Representative stress–strain curves under compression for all the examined aerogels, namely rGO, 90/10, 70/30 and 50/50. (b) Compressive modulus (left columns) and electrical conductivity (right columns) for all tested aerogel samples.

Both GAs and BNAs are highly efficient gas and oil absorbers24,43,44, thus we examine the capacity of the present aerogels for their use in gas absorption applications. We note that a potential application is the use of the proposed aerogels for the protection from volatile organic compounds (VOCs) and humidity of important artefacts in storage facilities (the number surpass the 10,000) which is one of the main objectives of the present work.

We tested the efficiency of the proposed hybrids and the GA for absorption of formaldehyde, hydrochloric acid and acetic acid, which are among the most commonly examined gases, as well the water vapour (humidity) absorption. Formaldehyde, in particular, is considered as a model polar VOC pollutant45. Prior to exposing the samples to the gas/humid environment, care was taken to dry the samples by heating to 200 °C, followed by immediate measurements of their initial weight under dry conditions. The samples were then placed in a desiccator and exposed to saturated gas controlled environment, while their mass weight gain due to absorption was monitored by regular gravimetric measurements until the maximum mass increment was reached. More details can be found in the experimental section. From Fig. 6 it is seen that as the concentration of hBN into the hybrid aerogels increases, so does the maximum absorption of formaldehyde. Thus, the most efficient absorber of formaldehyde is the hybrid aerogel with the highest amount of hBN, the 50–50 sample. We can observe that the 90–10, 70–30 and 50–50 samples show maximum absorptions higher by 3, 6 and 7 times in comparison to the neat rGO sample. This proves the capability of hBN in absorbing formaldehyde, and to the best of our knowledge, this is the first time that such a structure, like a hybrid aerogel from rGO and hBN, is reported to exhibit this behaviour. On the contrary, for the case of hydrochloric and acetic acid the opposite trend is observed with the rGO sample being the most efficient absorber. The absorption is slightly reduced for the 90–10 sample, while higher decrease occurs for the other cases.

Figure 6
figure 6

Absorption capacity of the rGO and hybrid rGO-hBN aerogels for formaldehyde, hydrochloric acid, humidity and acetic acid.

The high absorption of formaldehyde of the HAs can be mainly ascribed to the polarity of the molecules of formaldehyde. It is considered that materials with high specific surface area and a hydrophilic surface can be effective adsorbents for formaldehyde46. Formaldehyde is physiosorbed on graphene47,48, and this absorption is weak because of its small binding energy, large binding distance and small charge transfer from formaldehyde to the graphene sheets48. On the other hand, hBN (sponge like) exhibits superb formaldehyde absorption performance due to its high specific surface area and its abundant surface hydroxyls and amines46. Furthermore, the HAs are more hydrophilic compared to the rGO as deduced form the humidity absorption measurements discussed below, which also contribute to the high performance toward gaseous formaldehyde46. Additionally, chemisorption takes place on the hBN layers and contributes to high formaldehyde absorption via Cannizzaro-type disproportionation reactions, during which formaldehyde is converted into less toxic formic acid and methanol46. Another factor that can enhance the formaldehyde absorption is the surface properties of hBN. Hydrogen bonding between formaldehyde and hydroxyl or amine groups is very crucial in the absorption of formaldehyde49,50, while the formaldehyde molecules and hBN layers show both planar configuration, resulting in lower adsorption resistance46. Finally, formaldehyde has π-bonding between C and O atoms, while hBN also shows large 2D delocalized π-bonded structure. Consequently, π−π bonding interaction can occur between hBN and formaldehyde and this strong interaction enhances the absorption of the latter46.

As already mentioned, formaldehyde is known to be weakly physiosorbed on reduced graphene oxide sheets as a consequence of its small binding energy to graphene which is equal to Eb = 0.29 eV51. When the formaldehyde molecules are adsorbed onto the neat rGO sheets, formaldehyde shows the preference to be located perpendicular to the sheet with H atoms close to the sheet, as shown in Fig. 7a below. This results to an angle of 90° between the sheet and the axis passing through C = O bond of formaldehyde. Also, when formaldehyde is adsorbed onto the rGO sheets, creates an energy band gap at the Dirac point of graphene51,52. Thus, graphene then becomes a n-type semiconductor, the valence electrons are increased and charge transfer from the electron donor molecules, formaldehyde, to the graphene sheets takes place51.

Figure 7
figure 7

(a) Adsorption configuration of formaldehyde on reduced graphene oxide (rGO) sheet. (b) Possible disproportionation reaction for formaldehyde over the surface of porous boron nitride.

As the amount of hexagonal boron nitride is gradually increased in the hetero-structured aerogel, the adsorption of formaldehyde increases (from 90/10% wt. to 50/50% wt.). This can be explained due to the strong chemisorption between the molecules of (adsorbed) formaldehyde and the ‘doped’ with hBN, graphene. It has been shown that doping graphene with boron (B) and nitrogen (N) atoms enhances the interaction between the graphene and formaldehyde significantly53. Formaldehyde exhibits larger adsorption energy and net charge transfer when is chemisorbed on doped graphene in comparison to a corresponding undoped graphene sheet. Specifically, for a boron-doped graphene sheet, the adsorption energy (Eads) and charge transfer (Q) are equal to − 0.543 eV and 0.239 e respectively, while the corresponding values for pristine graphene are − 0.137 eV and 0.094 eV, respectively53. This high adsorption energy is a strong indication that chemical bonds are created between the dopants (B, N) and O atom of formaldehyde. The reason behind this behaviour is the hybridisation between the dopants (B, N) and C atoms, around the two dopant atoms. Especially for boron, an overlap between B-2p and C-2p orbitals, with the energy ranging from − 3.810 to − 2.236 eV is observed53, which proves the appearance of a strong hybridization between B and C atoms around the doping atom. This strong orbital interaction between the p orbitals of B, N, and the C atoms also alters the magnetic properties of such systems53.

For the interaction between the boron nitride platelets and the formaldehyde, Cannizzaro-type disproportionation reaction of the latter has been reported to be dominant54. The related mechanism is shown in Fig. 7b. During this reaction, the molecules of formaldehyde are adsorbed onto the boron nitride. The nitrogen atoms of the boron nitride sheet function as a Lewis base and initiate a reaction of nucleophilic addition during which a hydrogen atom of formaldehyde is transferred to another similar molecule (phase 1). Consequently, adsorbed formaldehyde molecules are transformed into HCONH2 (formamide) and others into methoxy salts (phase 2). Afterwards, the produced formamide and methoxy salts react with water adsorbed onto the surface of the material to produce in turn less toxic formic acid and methanol (phase 3). The last phase indicates the critical role of the hydrophilicity for the examined hybrid aerogel.

Regarding the adsorption of hydrochloric acid onto the examined pure rGO and hybrid aerogels, a weaker interaction between the HCl molecules and the surface of the prepared materials can be observed, as shown in Fig. 6. Hydrochloric acid is physiosorbed by graphene55,56 as can be deduced by the same approximately adsorption rates of all the hybrid aerogels that were examined in HCl, and neat rGO inside formaldehyde. It has been found that especially for a single chlorine atom, the bonding is ionic through the transfer of charge from graphene to chlorine adatom and creates a small local distortion in the underneath planar graphene55. The transfer of a single chlorine adatom on the surface of graphene is achieved almost without barrier. But, the anchoring of a graphene surface with Cl adatoms resulting in several conformations cannot be maintained because of the strong Cl-Cl interaction which in turn facilitates the desorption through the creation of Cl2 molecules.

The absorption capacity for acetic acid is a bit smaller as compared to the hydrochloric acid but a similar trend holds with decreasing absorption with increased amount of hBN. Similar experiments have been performed for testing reduced graphene oxide based porous materials as gas absorbers for various VOCs including benzene, toluene, carbon dioxide and formaldehyde57,58,59,60,61,62,63. For comparison they are listed in the supporting information (Table S1) including the results of the present work, along with their gravimetric measurements and details on the structure of the examined samples.

The capability of moisture absorption of the HAs was also examined after their insertion into a closed plastic chamber in which vials full of water were placed inside as the source of humidity. The relative humidity (RH) was measured to be ~ 97% by the addition of a humidity sensor inside the chamber. Following the same procedure with the gas absorption measurements, the samples were firstly dried at 200° C and then being weighed periodically until they achieved a maximum absorption. As shown in Fig. 6, the addition of hBN platelets increases significantly the humidity absorption which can be up to ~ 8 times higher than the absorption of the neat rGO sample. The increment reaches a value of ~ 1203.7% for the 50/50 HA sample of its initial weight. These results show that the hBN induces a highly hydrophilic nature to the HAs, in agreement with recent experiments on the wettability properties of pure hBN aerogels64. In Tables 3, 4, 5 we present the initial weight of the samples before subjecting them to the gas or humid environment, as well their weight after drying for removing the absorbed gas/humidity. The samples fully recover to their initial weight, exhibiting an excellent regeneration which a prerequisite for their effective use as gas absorbers. We also tested the influence of humidity in the absorption capacity for each examined VOC. We measured the humidity level inside each desiccator, and we performed independent measurements for humidity absorption at these RH levels. The RH inside the desiccators found to be in the range 63–75.5% with the upper value corresponding to the formaldehyde. From the independent RH measurements, we observed that the samples begin to gain weight due to humidity when the RH overcomes the value of RH ~ 70%. Close to the upper limit of RH ~ 75%, the corresponding percentage increase of the aerogel weight was found to be only 0.22% and it was lower for RH < 70%. Thus, we can safely deduce that the influence of humidity in the VOC absorption is negligible.

Table 3 Samples with formaldehyde.
Table 4 Samples with hydrochloric acid.
Table 5 Humidity absorption.

As already mentioned, we cannot fabricate pure hBN aerogels with this approach, thus we cannot have a direct comparison for this case. The absorption results show that the two materials have different efficiencies depending on the specific vapour. The absorption efficiency is not compromised by mixing of the 2D materials, thus, in order to make as efficient absorbers as possible for a wide range of vapour concentrations, we need to use more than one material. Given the large availability of 2D materials with diverse properties, there is much room for the fabrication of novel highly efficient absorbers beyond the state-of-the-art. Moreover, these hybrids can be exploited in other applications for which electrical and thermal conductivities are required.

Conclusions

In summary, we presented a facile fabrication route for 2D hybrid aerogels comprising of rGO mixed with hBN platelets. The fabrication method is cost effective and scalable compared to high energy demanding methods for the fabrication of pure hBN aerogels, and can be used for mixing other 2D materials with rGO. The resulted HAs are mechanically robust with mechanical properties similar to those of pure GA. The HAs are electrically conductive, despite a decrease that occurs compared to the pure rGO aerogel, which can be useful for applications for which electrical conductivity is required. The most striking result is the very effective absorption capacity of the HAs of formaldehyde and humidity, which is 7–8 times higher compared to the pure GA and at least one order of magnitude higher than a pure hBN aerogel46. Furthermore, the addition of hBN provides superior thermal conductivity to the HAs12, which make them promising materials for applications at elevated temperatures, further expanding their potential use. The HAs can be used as more efficient VOC absorbers for absorption of multiple compounds, in contrast to the standalone materials which are not efficient for all cases. All the above give the HAs a high degree of multi-functionality which enables their use in a plethora of applications, and, as demonstrated in the main text, the proposed aerogels can be used for protection of artefacts such as paintings in storage facilities in which a high concentration of humidity and VOCs is encountered.