Introduction

The pioneering study of graphene by Geim and Novoselov (Novoselov et al. 2004) made the world aware of a new member of carbon nanostructures. Since then, many interesting functional properties have been described (Neto et al. 2009; Zhang et al. 2011) together with unique bioactivity (Jastrzębska et al. 2012), drawing scientific attention and technological interest to graphene. Lately, it has been shown that graphene and its related materials show great promise for potential applications in, e.g. biosensors, adsorbents for heavy metal removal (Zhang et al. 2011), as well as lithium ion batteries and electrochemical supercapacitors (Li et al. 2011). It has also been shown that, by deposition of different nanoparticles on graphene sheets, it is possible to develop composite materials and, as a result, improve the properties of the deposited nanoparticles by utilizing the specific electrical properties of graphene (Jiang et al. 2011; Liang et al. 2010; Peter et al. 2015). It is also known that the modification of graphene oxide (GO) during in situ methods is realized mainly through covalent bonding between nanoparticle precursors and oxygen-containing functional groups that are known to exist on GO surface resulting in reduction to reduced graphene oxide (RGO) (Liang et al. 2010; Wang et al. 2010; Jastrzębska et al. 2015, 2016). The possibility of nanoparticle detachment from graphene surface is highly valuable (Singh et al. 2011). It is also accepted that the so-called wet sol–gel methods can be used for in situ GO modification techniques, e.g. the preparation of solid RGO/SiO2 thin films for transparent conductors (Watcharotone et al. 2007) as well as solid layers of RGO/Al2O3 for catalytic hydrodesulfurization (Tian et al. 2015). The latter method consists of the wet hydrolysis of boehmite in acidic environment and subsequent thermal treatment. The recent findings of Ziemkowska and co-workers (Ziemkowska et al. 2014) also indicate that the morphology of Al2O3 nanopowder synthesized using a wet sol–gel method is strictly dependent on the amount of water introduced to the reaction environment and that the lack of H2O results in obtaining agglomerates or sponge-like structures. Recently, a new dry sol–gel method for the synthesis of RGO(5 wt%)/Al2O3 nanocomposites has been investigated (Jastrzębska et al. 2015). The method was realized in a full-dry environment (no hydrolysis or condensation during the reaction), in contrast to conventional wet sol–gel methods (in which water is introduced to the reaction). The main disadvantage was that the Al2O3 nanoparticles often agglomerated on the surface of RGO flakes resulting in the undesirable aggregation of some part of the final product (not uniform, porous and agglomerated layer of Al2O3). Lately, a simplified sol–gel method has been analysed for the production of an RGO/Al2O3 (40 wt%) nanocomposite (Jastrzębska et al. 2016). The method was expected to be simplified relative to the dry sol–gel method. Mild hydrolysis and condensation were not separated as a single stage but were initiated at the start of the process as a result of H2O present in the air and occurred during the whole progress of the process. Aluminium triisopropoxide was used as an Al2O3 precursor, and dry isopropanol was the reaction environment. However, the agglomeration of Al2O3 nanoparticles on the edges and flexures of RGO was still observed. Al2O3 nanoparticles were distributed on the surface of GO flakes non-homogenously, and the measured average particle size was more than three times higher in comparison with free Al2O3 nanoparticles (Jastrzębska et al. 2016). The present paper gives new insight into the problem of controlling and optimizing the morphology of RGO/Al2O3 nanocomposites. The dry and simplified sol–gel methods for the synthesis of RGO/Al2O3 nanocomposites were also compared, and the influence of six key synthesis parameters on the morphology of the resulting nanocomposite powders was investigated to reduce the undesired agglomeration of the nanocomposite flakes to a significant minority. This is also the first study with a detailed discussion of the reactions expected to occur during the synthesis of an RGO/Al2O3 nanocomposite together with the proposed reaction sequences.

Experimental

Reduction of GO to RGO

In the present study, we used commercial GO flakes (Cheap Tubes Inc., USA) and we prepared RGO flakes by ourselves. The vacuum-assisted thermal reduction process was used for the reduction of GO to RGO. It is accepted that thermal reduction is a simple and clean way to almost completely remove oxide-containing functional groups from GO surfaces by utilizing vacuum-assisted heat treatment (Erickson et al. 2010; Mavarro et al. 2010; Yang et al. 2009a; Mattevi et al. 2009). Our results show that the thermal treatment of GO flakes at 150 °C in a 3 × 10−2 Torr vacuum for a period of 6 h resulted in a change of the colour of GO from brown to black. The GO and RGO flakes were further used in the reaction processes.

Different parameters of RGO/Al2O3 nanocomposite synthesis

It is known that hydrophobic/hydrophilic incompatibility between graphene and metal oxide nanoparticles makes their deposition highly difficult (Wang et al. 2008). Conversely, GO is also widely known as a good candidate for supporting metal or metal oxide nanoparticles due to the presence of reactive oxygen-containing species on its surface (C–OH, O=C–OH, C=O, or C2O). In this study, the compatibility between RGO and Al2O3 nanoparticles is addressed in detail in parallel to the optimized method of its synthesis. The considered synthesis processes leading to different RGO/Al2O3 nanocomposite powders with different morphologies are presented in Fig. 1. As can be seen, the influence of six key synthesis parameters on the morphology of resulting RGO/Al2O3 nanocomposite powders was investigated. The analysis took into account parameters such as:

Fig. 1
figure 1

Flowchart presenting different procedures of RGO/Al2O3 nanocomposite synthesis considered for the analysis of the influence of key synthesis parameters on the composite morphology

Full size image
  1. 1.

    the type of aluminium oxide organic precursor (triethylaluminium, Et3Al, or aluminium triisopropoxide, Al(i-PrO)3);

  2. 2.

    the type of reaction environment (hexane or isopropanol (i-PrOH));

  3. 3.

    the type of graphene used as a substrate for the deposition of Al2O3 nanoparticles (GO or RGO);

  4. 4.

    the thermal annealing temperature (280 or 500 °C);

  5. 5.

    the concentration of RGO(0, 5, 10, or 15 wt%) in the final product;

  6. 6.

    the concentration of the modifier, diethyl ether (Et2O) (1 or 3 wt%).

The procedures of sol–gel synthesis of RGO/Al2O3 nanocomposites were as follows:

For the R1 process, 500 ml of dry hexane was introduced to a 1000 ml reactor equipped with a magnetic stirrer. Then, 0.01 g of RGO flakes was added to the hexane, and the resulting suspension was stirred for 1 h at 1000 rpm. Subsequently, 0.5 ml of 1 wt% solution of trimethylaluminium was introduced, and the reaction mixture was stirred in air until the spontaneous evaporation of the solvent. After that, the dark brown powder of the intermediate product (organic precursor) was obtained. After the thermal decomposition of the precursor (280 °C, 3 h, air atmosphere) in a muffle furnace, the black final product of the RGO(5 wt%)/Al2O3 nanocomposite was obtained.

For the R2-4 processes 500 ml of dry hexane was introduced to a 1000 ml reactor equipped with a magnetic stirrer. Then, 0.1 g of GO flakes was added to the hexane, and the resulting suspension was subsequently stirred for 1 h at 1000 rpm. Subsequently, 0.5 ml of 1 wt% solution of trimethylaluminium was introduced, and the reaction mixture was stirred in air until the spontaneous evaporation of the solvent. After that, the dark brown powder of the intermediate product (organic precursor) was obtained. The sample of the organic precursor was divided into three parts. The first part of the sample was directly subjected to SEM analysis (see the R2 process). The second part of the sample was subjected to thermal decomposition at 500 °C for 3 h in air atmosphere in a muffle furnace (see the R3 process), whereas the third part of the sample was thermally decomposed at 280 °C for 3 h in air atmosphere (see the R4 process). After thermal annealing, the black final products of the RGO(5 wt%)/Al2O3 nanocomposites were obtained.

For the R5 and R6 processes 500 ml of dry hexane was introduced to a 1000 ml reactor equipped with a magnetic stirrer. Then, 0.01 g of GO flakes was added to the hexane, and the resulting suspension was subsequently stirred for 1 h at 1000 rpm. The reaction mixture was divided into two parts, and 1 and 1.5 ml of 2 or 3 wt% solution of trimethylaluminium was introduced for the R5 and the R6 processes, respectively. Then, the reaction mixtures were stirred in air until the spontaneous evaporation of the solvent. After that, the dark brown powder of the intermediate products (organic precursors) was obtained. After thermal annealing at 280 °C for 3 h in air atmosphere, the black final products of the RGO(10 wt%)/Al2O3 and RGO(15 wt%)/Al2O3 nanocomposites were obtained for the R5 and R6 processes, respectively.

For the R7 and R8 processes 500 ml of dry hexane was introduced to a 1000 ml reactor equipped with a magnetic stirrer. Then, 0.01 g of GO flakes was added to the hexane, and the resulting suspension was subsequently stirred for 1 h at 1000 rpm. The reaction mixture was divided into two parts, and 0.5 and 1.5 ml of 1 or 3 wt% solution of trimethylaluminium was introduced for the R7 and the R8 processes, respectively. Then, the reaction mixtures were stirred for 0.5 h at 1000 rpm. After that time, 0.003 ml and 0.01 ml of 8.65 × 10−5 or 2.58 × 10−5 mol of (Et)2O was added for the R7 and R8 processes, respectively. Subsequently, the reaction mixtures were stirred for 0.5 h at 1000 rpm. After that time, the reaction mixtures were stirred in air until the spontaneous evaporation of the solvent. The dark brown powder of the intermediate products (organic precursors) was obtained. After thermal annealing at 280 °C for 3 h in air atmosphere, the black final products of the RGO(5 wt%)/Al2O3 and RGO(15 wt%)/Al2O3 nanocomposites were obtained for the R7 and R8 processes, respectively.

For the R9 process 500 ml of dry isopropyl alcohol was introduced to a 1000 ml reactor equipped with a magnetic stirrer. Then, 0.01 g of GO flakes was added to the isopropyl alcohol, and the resulting suspension was subsequently stirred for 1 h at 1000 rpm. Then, 0.02 g of aluminium triisopropoxide was introduced. After that, the reaction mixture was stirred in air until the spontaneous evaporation of the solvent, and the dark brown powder of the intermediate product (organic precursor) was obtained. After thermal annealing at 280 °C for 3 h in air atmosphere, the black final product of the RGO(5 wt%)/Al2O3 nanocomposite was obtained.

For the synthesis of Al2O3 nanopowder, we used a process similar to the one presented for the R4 process, excluding the stage of GO addition.

Chemicals and substrates used

Chemicals used in the RGO/Al2O3 nanocomposite synthesis (hexane, isopropyl alcohol, triethylaluminium, aluminium triisopropoxide, diethyl ether) were purchased from Sigma-Aldrich (Gillingham, Dorset, UK). Trimethylaluminium was moisture sensitive, and it was stored under argon atmosphere. GO flakes were purchased from Cheap Tubes Inc. (USA) and were synthesized from natural graphite powders by a modified Hummer’s method, which uses a combination of potassium permanganate and sulphuric acid.

Investigation methodology

The GO flakes, RGO flakes, RGO/Al2O3 nanocomposites, and Al2O3 nanopowder were examined in terms of the obtained morphology and physical properties. The methodology used for the characterization of the obtained materials was similar to that described in our previous publications (Jastrzębska et al. 2015, 2016). Briefly, a scanning electron microscope (SEM-LEO 1530, Zeiss, USA) was used for morphology observations. Samples were deposited onto the surface of a carbon tape and coated with a thin carbon layer The elemental EDS analysis was used for the analysis of the local contents of the elements present in a given region for the chosen RGO/Al2O3 nanocomposite. The obtained SEM images were subsequently analysed quantitatively using statistical (stereological) analysis and the MicroMeter v.086b computer program. The analysis took into account the mean value E(x) and the coefficient of variation CV(x) of the stereological parameters such as the equivalent diameter d 2; the maximum value of the chord of particle d max; the minimum value of the chord of particle d min; the particle perimeter p; the Couchi perimeter pC, as well as the particular coefficients of shape: d max /d 2 , d min /d 2 , p/πd 2, and p/pC. The additional parameter n represents the number of measured values.

The isotherms of physical nitrogen sorption V = f(p/p°) were measured experimentally, using a Quadrasorb-SI device (Quantachrome Instruments, USA) at a temperature of −195.8 °C, within the entire range of the relative pressure. The specific surface area, S BET, was determined using the Brunauer, Emmett, Teller (BET) method and the total volume of the pores, V BJH, present in the samples was determined using the Barret, Joyner, Halenda (BJH) method. The distributions of pore sizes were further used for the estimation of the average pore size, D BJH.

The quantitative chemical composition of the RGO/Al2O3 nanocomposite was also analysed using the XPS spectroscopy (Microlab 350 spectrometer from Thermo Electron) with soft X-Ray radiation as the source of excitation (AlKα radiation, hν = 1486.6 eV; Al; 200 W). The peak of C1s = 285.0 eV (binding of CC/CH) was used as the internal standard.

Results and discussion

Recently, we have published the preliminary investigations results on a new dry sol–gel method (Jastrzębska et al. 2015) and a simplified sol–gel method (Jastrzębska et al. 2016) of the synthesis of well-dispersible RGO/Al2O3 nanocomposites. To obtain RGO flakes, the vacuum-assisted thermal reduction process was used in the present study for the reduction of GO. It is known that to obtain RGO two possible chemical routes can be applied in particular: methods where a solvent is used (wet methods) and the so-called dry methods. Numerous literature data have been published on the wet chemical reduction of GO sheets. It can be performed with several reducing agents, e.g. hydrazine (Becerril et al. 2008), sodium borohydrate (Shin et al. 2009), hydroquinone (Wang et al. 2008), or even strongly alkaline solutions (Fan et al. 2008). During the reduction process, the brown-coloured dispersion of GO in the solvent turns black and the reduced sheets aggregate and precipitate (Stankovich et al. 2007). Dry reduction methods are, however, cleaner than wet ones and can be performed using, e.g. gaseous hydrogen (after thermal expansion) (Wu et al. 2009). Thermal reduction is a simple and clean way to reduce GO and utilizes heat treatment to remove the oxide functional groups from GO surfaces. Different heat treatments have been used so far at different final temperatures (200, 500, and 1000 °C) in argon atmosphere or ultrahigh vacuum (10−8 Torr) to study the possible reduction of films composed of GO (Yang et al. 2009b). A significant reduction of GO has been observed, resulting in the complete disappearance of C=O and O=C–OH species with the exception of C–OH. The authors have suggested that, even after the 1000 °C treatment of GO, the remaining C–OH groups are still present because their reduction is very difficult from the thermodynamical point of view. Other authors’ results also confirm this assumption (Erickson et al. 2010; Mavarro et al. 2010). It is accepted that even ultrahigh vacuum assistance (Mattevi et al. 2009) is not sufficient to completely remove all of the oxygen from the GO surface. Our observations showed that the heat treatment of GO flakes at 150 °C in vacuum for a period of 6 h resulted in the colour of GO changing from brown to black.

Morphology

The morphology of GO flakes and RGO flakes (used as substrates for Al2O3 nanoparticle deposition) was examined qualitatively using a scanning electron microscope (SEM). Figure 2 shows the morphology of the analysed GO and RGO samples. It can be seen that both GO and RGO flakes possessed an irregular shape which can be described as ‘curly’ (Fig. 2a, b, respectively). Our results indicate that thermal annealing made it possible to obtain a similar morphology of the resulting RGO flakes in comparison with GO. Also, the high magnification views show the smooth surfaces of GO and RGO flakes. Their lateral dimensions were ca. 4 µm.

Fig. 2
figure 2

SEM images obtained for a the GO and b the RGO flakes

Full size image

The influence of six key synthesis parameters on the morphology of the resulting RGO/Al2O3 nanocomposite powders was investigated qualitatively using an SEM. The analysis took into account parameters, such as: (1) type of aluminium oxide organic precursor (trimethylaluminium, Et3Al, or aluminium triisopropoxide, Al(i-PrO)3); (2) type of reaction environment (hexane or isopropanol (i-PrOH)), (3) type of graphene used as a substrate for the deposition of Al2O3 nanoparticles (GO or RGO), (4) thermal annealing temperature (280 or 500 °C), (5) concentration of RGO(0, 5, 10, or 15 wt%) in the final product, and (6) concentration of the modifier, Et2O (1 or 3 wt%). The obtained investigation results showed that the above-mentioned factors have a major influence on the morphology of the investigated RGO/Al2O3 nanocomposites.

The morphologies of RGO(5 wt%)/Al2O3 nanocomposites obtained in the R4 and R9 processes are presented in Fig. 3a, b, respectively. The obtained nanocomposites differed only in the type of aluminium oxide organic precursor used in the synthesis process. Trimethylaluminium (Et3Al) was used as the Al2O3 organic precursor in the R4 process, whereas aluminium triisopropoxide (Al(i-PrO)3) was used in the R9 process. It can be seen that the RGO(5 wt%)/Al2O3 nanocomposite obtained by the R4 process was characterized by less agglomeration in comparison with the nanocomposite obtained in the R9 process. The Al2O3 nanoparticles deposited on the surface of RGO flakes (Fig. 3a) were much easier to identify and, as can be seen in the high magnification image, acted as a spreader between individual RGO flakes. The agglomeration of the RGO(5 wt%)/Al2O3 nanocomposite obtained by the R9 process (Fig. 3b) was so high that not even a single RGO flake was able to be identified. It should also be noted that hexane was used as the reaction environment in the R4 process, whereas isopropanol (i-PrOH) was used in the R9 process. As can be seen, the best morphology in the case of the RGO(5 wt%)/Al2O3 nanocomposite was obtained when trimethylaluminium was used as an Al2O3 nanoparticle organic precursor together with hexane as the reaction environment.

Fig. 3
figure 3

SEM images obtained for the RGO(5 wt%)/Al2O3 nanocomposites obtained in a the R4, b the R9, and c the R1 process

Full size image

The morphologies of RGO(5 wt%)/Al2O3 nanocomposites obtained in the R4 and R1 processes are presented in Fig. 4a, c, respectively. The obtained nanocomposites differed in the type of graphene used as the substrate for the deposition of Al2O3 nanoparticles. In the R4 process, GO was used as the substrate, whereas in the R1 process, RGO was used. As can be seen on SEM images, GO was a better choice for a substrate. RGO(5 wt%)/Al2O3 nanocomposites obtained using RGO were characterized by higher agglomeration.

Fig. 4
figure 4

SEM images obtained for the RGO(5 wt%)/Al2O3 nanocomposites thermally decomposed at a 0 °C, b 280 °C, and c 500 °C

Full size image

The morphologies of RGO(5 wt%)/Al2O3 nanocomposites obtained in the R2, R4, and R3 processes are presented in Fig. 4a–c, respectively. The obtained nanocomposites differed in the temperature of thermal annealing used in the synthesis process. The powder obtained in the R2 process (RGO/Al2O3 organic precursor) was subjected to SEM observations without any thermal treatment, directly after the evaporation of the solvent. The RGO(5 wt%)/Al2O3 nanocomposite obtained in the R4 process was thermally decomposed at 280 °C, whereas 500 °C was used in the R3 process for the thermal decomposition of the organic precursor. The investigation results indicate that the process of thermal treatment has a significant influence on the morphology of the final product. The RGO(5 wt%)/Al2O3 nanocomposite thermally treated at 500 °C was much more agglomerated than the one treated at 280 °C. As can be seen in Fig. 4, thermal treatment generally stimulates the growth and agglomeration of Al2O3 nanoparticles deposited on the edges of RGO flakes. When considering the organic precursor, it can be seen that the number of Al2O3 agglomerates attached to the edges of RGO flakes was a minority and also a majority in RGO(5 wt%)/Al2O3 nanocomposite thermally treated at 500 °C. This indicates that the temperature chosen for the thermal decomposition of the nanocomposite organic precursor should be far below 500 °C. Thus, the temperature of 280 °C should be considered optimal for the decomposition of aluminium triethoxide and the formation of Al2O3 nanoparticles characterized most probably by a low-temperature (amorphous) phase of alumina.

The morphologies of RGO/Al2O3 nanocomposites with 0, 5, 10, and 15 wt% of Al2O3 addition are presented in Fig. 5. The above-mentioned concentrations of Al2O3 in the final product correspond to the each of the particular reaction processes, such as: R4 (Al2O3), R4, R5, and R6, respectively. For the synthesis of the Al2O3 nanopowder, we used a similar process to that described for R4, excluding the stage of GO addition. The morphology of the Al2O3 nanopowder is presented in Fig. 5a. As can be seen, it is composed of agglomerates and single nanoparticles of sizes not exceeding 200 nm. This morphology can be described as typical for Al2O3 nanopowders obtained by us previously in sol–gel processes (Jastrzebska et al. 2011a, b; c, 2013, 2014, 2015b; Ziemkowska et al. 2014; Polis et al. 2013; Jusza et al. 2011, Karwowska et al. 2002). In the R4 process, the 5 wt% addition of RGO flakes to the nanocomposite results in less agglomeration of the final product as a result of the deposition of Al2O3 nanoparticles on the surface of RGO flakes (Fig. 5b). The rest of the alumina organic precursor, not utilized on the surface of the flakes, existed as agglomerates composed mainly of Al2O3 nanoparticles. As can be seen, the reduction of Al2O3 sizes when deposited on RGO was achieved. It should be mentioned that in our previous investigation results the average particle size of Al2O3 deposited on RGO was more than three times higher in comparison with free Al2O3 nanoparticles (Jastrzębska et al. 2016). It is expected that an increase of the RGO addition should decrease the presence of free Al2O3 agglomerates. This was, however, not confirmed with the RGO/Al2O3 (10 wt%) sample because of final product agglomeration (Fig. 5c). On the other hand, when considering the nanocomposite with 15 wt% addition of RGO flakes (Fig. 5d), it can be seen that the number of free Al2O3 agglomerates highly decreased in comparison with the rest of the samples. As can be seen, however, on a high magnification image, some of the GO flakes were not covered with Al2O3 nanoparticles.

Fig. 5
figure 5

SEM images obtained for the RGO/Al2O3 nanocomposites with a 0 wt%, b 5 wt%, c 10 wt%, and d 15 wt% RGO addition. The samples were thermally decomposed at 280 °C

Full size image

The morphologies of RGO(5 wt%)/Al2O3 nanocomposites obtained in the R4 and R7 processes are presented in Fig. 6. The nanocomposite obtained in the R7 process was modified with 0.006 ml of Et2O. As compared with the RGO(5 wt%)/Al2O3 nanocomposite non-modified with Et2O (Fig. 5a), the above-mentioned modification resulted in the higher agglomeration of the final product and the formation of a high amount of Al2O3 agglomerates attached to the edges and the surface of RGO flakes (Fig. 6b).

Fig. 6
figure 6

SEM images obtained for the RGO/Al2O3(5 wt%) nanocomposite modified with a 0 ml and b 0.006 ml Et2O

Full size image

The morphologies of RGO(15 wt%)/Al2O3 nanocomposites obtained in the R6 and R8 processes are presented in Fig. 7. The nanocomposite obtained in the R8 process was modified with 0.01 ml of Et2O. As compared with the RGO(15 wt%)/Al2O3 nanocomposite non-modified with Et2O (Fig. 7a), the above-mentioned modification resulted in improvement in covering the surface of RGO flakes (Fig. 7b). This can be clearly seen in the high magnification images of single RGO flakes. However, the low magnification images revealed the undesirable agglomeration of the final product.

Fig. 7
figure 7

SEM images obtained for the RGO/Al2O3(15 wt%) nanocomposite modified with a 0 and 0.01 ml Et2O

Full size image

The representative SEM–EDS pattern of the RGO/Al2O3 nanocomposite with 5 wt% RGO addition is shown in Fig. 8. As can be observed, EDS analysis taken from the surface of an RGO flake with deposited Al2O3 nanoparticles (sharp-marked point) displayed strong C, O, and S peaks next to Al and Na peaks. It should be noted that the existence of S and Na peaks corresponded to the presence of impurities in GO (as a result of using the specific reagents in the modified Hummer’s method). The quantitative results of the EDS analysis are presented in Table 1. They reveal the presence of large amounts of carbon (over 80 wt%) and oxygen (ca. 16 wt%) together with a very low amount of aluminium (below 1 wt%). It should, however, be noted that the coverage of the RGO surface by Al2O3 nanoparticles is generally not uniform and the amount of single nanoparticles on the surface and edges of RGO flakes is different. The analysis was taken from the point specifically chosen to be relative to the uniform coverage of the RGO surface. Thus, the obtained results indicate that the amount of Al2O3 nanoparticles that are able to create the uniform layer on the surface of RGO flakes is de facto not very high.

Fig. 8
figure 8

SEM image of a the RGO/Al2O3(5 wt%) nanocomposite and b the EDS analysis taken from the sharp-marked point of the analysed sample

Full size image
Table 1 Quantitative results of the EDS analysis of the RGO/Al2O3 (5 wt%) nanocomposite
Full size table

The microscopic examinations were also aimed at quantitatively characterizing the morphology of Al2O3 nanoparticles embedded on the surface of RGO flakes as well as making a comparison with the Al2O3 nanopowder using a statistical (stereological) analysis. The quantitative analysis was performed on the RGO/Al2O3(5 wt%) nanocomposite obtained in the R4 process, since it was the most efficient way to prepare the nanocomposite with the best morphology. The results of the particle size analysis are shown in Fig. 9a and in Table 2. As can be seen in Table 2, there were no changes in the shape of Al2O3 nanoparticles deposited on the surface of RGO flakes and free Al2O3 nanoparticles. The coefficients of shape were close to 1. However, a high reduction of the average value of particle size (d 2) was obtained for Al2O3 nanoparticles when deposited on the surface of RGO flakes. The sizes of Al2O3 particles deposited on RGO ranged from ca. 20 to 140 nm (the average value being 58 nm) and were lower in comparison with free Al2O3 nanoparticles, characterized by sizes ranging from ca. 20 to 180 nm (the average value being 95 nm). It should also be noted that free Al2O3 nanoparticles had a high agglomeration tendency. The average agglomerate size measured for Al2O3 nanopowder was over 0.3 µm (Fig. 9c) with a relatively low value of the coefficient of variation (0.14) (Table 3).

Fig. 9
figure 9

The distribution of the Al2O3 particles size deposited a on the surface of RGO flakes as well as b on free Al2O3 nanoparticles and c on agglomerates of the Al2O3 nanopowder comparative sample obtained in the R4 process

Full size image
Table 2 Results of the particle stereological analysis obtained for Al2O3 particles deposited on the surface of RGO flakes and free Al2O3 particles of the Al2O3 nanopowder comparative sample obtained in the R4 process
Full size table
Table 3 Results of the agglomerate stereological analysis obtained for the Al2O3 nanopowder comparative sample obtained in the R4 process
Full size table

Physical properties

The results of the investigations of the physical properties of GO and the obtained RGO are presented in Table 4, which shows the values of the specific surface area, S BET, the total pore volume, V BJH, and the average pore diameter, D BJH, of the analysed samples. As expected, the specific surface area of RGO flakes was 1.2 times higher in comparison to GO flakes (Table 4). It is accepted that the presence of chemisorbed water and surface oxygen species generally decreases the surface area measured by physical nitrogen sorption. Thus, our results of increase in the RGO flake BET surface area confirmed that the thermal treatment of GO resulted in the removal of oxygen-containing species from the surface of GO flakes. The specific surface area of Al2O3 nanopowder was much lower (53 m2/g) in comparison with materials containing graphene. GO and RGO flakes were characterized by similar values of total pore volume, which exceeded 0.1 cm3/g. It should be noted that the pore volume of the RGO/Al2O3 (5 wt%) nanocomposite, obtained in the R4 process, was very high (ca. 0.7 cm3/g) in comparison with other samples.

Table 4 Physical properties of GO and RGO flakes as well as the RGO(5 wt%)/Al2O3 nanocomposite powder obtained in the R4 process
Full size table

The nitrogen sorption isotherms obtained for the analysed samples are presented in Fig. 10. The obtained results are in good agreement with the SEM investigation results. They indicate that the shape of the isotherm obtained for RGO and GO flakes was typical for graphenes, characterized by the presence of slit-shaped pores (Srinivas et al. 2010). For the RGO/Al2O3 nanocomposite, the shape of the isotherm suggested the presence of bottle-shaped pores. It was hard to compare the shape of the isotherm obtained for pure Al2O3 due to the very low N2 adsorption; however; it is already accepted that the shape of the pores of Al2O3 is generally spherical or/and bottled (Qadeer and Ikram 2005). Taking into consideration that N2 particles only adsorb on available surfaces, it can be concluded that the pores in our RGO/Al2O3 nanocomposite originated mainly from Al2O3 nanoparticles.

Fig. 10
figure 10

Nitrogen sorption isotherms obtained for GO and RGO flakes, the RGO(5 wt%)/Al2O3 nanocomposite, as well as free Al2O3 particles

Full size image

XPS spectroscopy

XPS spectroscopy was used for the analysis of the RGO(5 wt%)/Al2O3 nanocomposite synthesized in the R4 process. The XPS analysis showed the quantitative chemical composition in the sample as well as the identification of the chemical state of elements present (Table 5). The C1 s peak located at 285 eV was assigned to the presence of the C–C bonds of graphitic carbon (17 at %). The other C1 s peaks were a result of the presence of –C–OH and O=C–OH oxygen-containing species, also according to other researchers (Yang et al. 2009b). The atomic percentage of –C–OH and O=C–OH bonds in the RGO/Al2O3 nanocomposite was ca. 7 and 2 at%, respectively. Comparing our results with the typical amount of oxygen on GO (ca. 50 wt%, Jastrzębska et al. 2016), one can observe a considerable reduction of the amount of oxygen groups and the reduction of GO to RGO as a result of the reaction process with triethylaluminium. The Al2p peaks at 74.4 eV (25.21 at %) and the O1 s peaks at 531.2 eV (17.24 at %) and 532.6 eV (21.58 at %) were assigned to the presence of crystalline Al2O3 on the surface of RGO. It should be noted that the presence of the crystalline phase of Al2O3 was a result of the thermal treatment of the obtained organic precursor. The XPS spectra also revealed the presence of carbonate (CO3 2−) species, in the corresponding O1s peak and C1s peak located at 290.8 and 529.4 eV, respectively. The obtained results thus suggest that the thermal decomposition was an incomplete process, and some amount of the organic precursor remained in the final product. This is justified by taking into consideration that the temperature of 280 °C may not be enough for an aluminium triethoxide to fully decompose to aluminium oxide. However, it should be noted that the amount of these species was not considerably high, 1.10 and 3.40 at %, respectively.

Table 5 The chemical state of elements present in the RGO(5 wt%)/Al2O3 nanocomposite obtained in the R4 process
Full size table

Discussion on reactions

We also studied the reactions that may occur at the interface between GO and triethyl aluminium, Et3Al (or (C2H5)3Al) on basis of the obtained results. The sol–gel reactions that occurred during RGO/Al2O3 nanocomposite synthesis are represented by the following formulae:

$${\text{RGO}} + ({\text{C}}_{2} {\text{H}}_{5} )_{3} {\text{Al}}\mathop{\longrightarrow}^{{{\text{O}}_{2} ,\;{\text{C}}_{6} {\text{H}}_{14} }}({\text{C}}_{2} {\text{H}}_{5} - {\text{O}})_{2} {\text{Al}} - {\text{O}} - {\text{GR}} + {\text{C}}_{2} {\text{H}}_{5} {\text{OH}}$$
(1)
$$({\text{C}}_{2} {\text{H}}_{5} - {\text{O}})_{2} {\text{Al}} - {\text{O}} - {\text{GR}}\mathop{\longrightarrow}^{{{\text{H}}_{2} {\text{O}},\;{\text{drying}}}}({\text{C}}_{2} {\text{H}}_{5} )_{2} {\text{AlOH}} - {\text{GR}}$$
(2)
$$({\text{C}}_{2} {\text{H}}_{5} - {\text{O}})_{2} {\text{Al}} - {\text{O}} - {\text{GR}}\mathop{\longrightarrow}^{{{\text{H}}_{2} {\text{O}},\;{\text{drying}}}}({\text{C}}_{2} {\text{H}}_{5} )_{2} {\text{Al}} - {\text{O}} - {\text{Al}}({\text{C}}_{2} {\text{H}}_{5} )_{2}$$
(3)
$$({\text{C}}_{2} {\text{H}}_{5} - {\text{O}})_{2} {\text{Al}} - {\text{O}} - {\text{GR}}\mathop{\longrightarrow}^{{{\text{H}}_{2} {\text{O}},\;{\text{drying}}}}{\text{C}}_{2} {\text{H}}_{5} {\text{Al}} - {\text{O}}$$
(4)
$$({\text{C}}_{2} {\text{H}}_{5} )_{2} {\text{AlOH}} - {\text{GR}} + ({\text{C}}_{2} {\text{H}}_{5} )_{2} {\text{Al}} - {\text{O}} - {\text{Al}}({\text{C}}_{2} {\text{H}}_{5} )_{2} - {\text{GR}} + {\text{C}}_{2} {\text{H}}_{5} - {\text{Al}} - {\text{O}} - {\text{GR}}\mathop{\longrightarrow}^{{{\text{O}}_{2} ,\;280{\kern 1pt}^\circ {\text{C}},\;3{\kern 1pt} {\text{h}}}}{\text{Al}}_{2} {\text{O}}_{3} - {\text{GR}} + {\text{CO}}_{2} \uparrow + {\text{H}}_{2} {\text{O}} \uparrow + {\text{C}}$$
(5)

The detailed reaction sequences are presented in Fig. 11. In the scheme of the reaction sequence A, an influence of water is exhibited, while the reaction sequence B presents reactions with atmospheric dioxygen. It should be noted that all of the reactions that may occur are already known in the subject literature. According to Ziemkowska and co-workers (Ziemkowska et al. 2014), aluminium trialkyls undergo oxidation, hydrolysis, or both oxidation and hydrolysis depending on the time of triethyl aluminium exposure to air. When H2O and O2 are present deficiently in the reaction, Al2O3 nanoparticles should be formed and a mild hydrolysis–oxidation-association process should occur because the reaction mixture is just exposed to air (like in the R4 process).

Fig. 11
figure 11

Scheme describing probable sol–gel reactions (divided into the two sequences) occurring during the RGO/Al2O3 nanocomposite synthesis using the R4 process. The ‘Et’ symbol represents the ethyl (C2H5) group

Full size image

During the production of RGO/Al2O3 nanocomposite, Et3Al was added in an excess to the number of functional groups containing oxygen atoms and bonded to the surface of RGO. According to the literature data (Pasynkiewicz et al. 1977; Bolesławski et al. 1983a, b; Storre et al. 1996; Mason et al. 1993), the first step of an interaction of Et3Al with these functional groups is a formation of unstable coordinate complexes (product A, Fig. 2) with coordinate bonds of Al←OH. The complexes undergo a further reaction with an evolution of ethane and OAlEt2 moieties formation (product B). Only the complex of Et3Al with the epoxide oxygen is stable. The reaction of the product B with water originating from moisture is the next step, in which ethyl groups bonded to aluminium atoms are gradually removed and substituted by OH groups. Et3Al molecules coordinated to the epoxide oxygens can be eliminated in the form of ethylhydroxy aluminium Al(OH)3−n Et n . In the obtained product C, the tri-coordinate aluminium atoms have a tendency to increase their coordination number by association with Al(OH)3−n Et n molecules, mainly formed in the reaction of free Et3Al with moisture.

As a result of an interaction of Al(OH)3−n Et n with tri-coordinate aluminium atoms, four-membered Al2O2 cycles are formed that are typical for aluminium alkoxides and hydroxides. The thus obtained product D is finally subjected to thermal treatment to remove water and organic moieties in the form of alkanes. The final RGO/Al2O3 nanocomposite contains aluminium oxide bonded to the surface of RGO.

Dioxygen present in the air has also an influence on RGO/Al2O3 nanocomposite formation (pathway B). According to the literature data (Barron 1993; Lewiński et al. 1996, 1999, 2000), an oxidation of aluminium alkyls leads to a mixture of aluminium alkoxides. In the product B1 (Fig. 11), an interaction of dioxygen with ethyl groups bonded to aluminium atoms results in the transformation of EtAl groups into EtOAl moieties and the formation of the product C1. The product C1 has a chance to react with water to eliminate ethanol and aluminium hydroxide and yield the product D1. After association with Al(OH)3−n Et n molecules present in the solution yielding the product D1 is transformed into E1 containing tetra-coordinate aluminium atoms. Similarly to the reaction sequence A, the final step of the RGO/Al2O3 nanocomposite production is thermal treatment of the precursor. Because there is an excess of triethyl aluminium in solution, extra molecules of Et3Al and Al(OH)3−n Et n associate with the EtAl AlOH and AOEt groups bonded with RGO. Therefore, after thermal decomposition of precursors (products D and E1), much more than one aluminium atom per a functional hydroxyl, carboxyl or epoxide group is present in the nanocomposite. These aluminium atoms together with oxygen atoms form a network chemically bonded to the surface of RGO.

Conclusions

The present paper gives new insight into the problem of controlling the morphology of reduced graphene oxide/alumina (RGO/Al2O3) nanocomposites. The dry and simplified sol–gel methods of RGO/Al2O3 nanocomposite synthesis were compared and the influence of six key synthesis parameters on the morphology of the resulting nanocomposite powders was investigated to optimize the morphology of RGO/Al2O3 nanocomposites in terms of reducing the undesired agglomeration of RGO/Al2O3 nanocomposite flakes to a significant minority and obtaining the uniform coverage of the RGO surface with Al2O3 nanoparticles. Our investigations confirmed that the synthesis parameters have a major influence on the morphology of the investigated RGO/Al2O3 nanocomposites. The obtained results indicate that the aluminium triisopropoxide, Al(i-PrO)3, makes it possible to obtain a more uniform morphology of the RGO/Al2O3 nanocomposite in comparison with triethylaluminium, Et3Al. It is also better to use hexane as the reaction environment and GO as a substrate for Al2O3 deposition in contrast to isopropanol (i-PrOH) and RGO, respectively. Diethyl ether used as a modifying agent enabled better and more uniform coverage of GO flakes but also resulted in the undesirable agglomeration of the final product. Also, despite a high excess of Al2O3 used (95 wt%), the lowest RGO/Al2O3 flake agglomeration and the formation of an uniform layer composed of Al2O3 nanoparticles were obtained only when 5 wt% of GO was used in the process. The optimal temperature for thermal decomposition was 280 °C for 3 h under an air atmosphere. The optimal method for RGO/Al2O3 nanocomposite synthesis was the R4 process. The nanocomposite thus obtained was characterized by the average size of surface Al2O3 particles of 58 nm, the BET-specific surface area of 242.4 m2/g, and open porosity of 0.7 cm3/g.

Our results indicate that the optimization of the synthesis process enabled the reduction of the undesired agglomeration of the final product to an insignificant minority. The average size of the Al2O3 nanoparticles deposited on the surface of RGO was reduced by ca. 50% in comparison with free Al2O3 nanoparticles. Our results also show that during other reaction pathways Al2O3 nanoparticles cover RGO flakes non-uniformly. As expected, thermal treatment stimulated the growth and agglomeration of Al2O3 nanoparticles deposited on the edges of RGO flakes, which cannot be generally avoided but can be controlled and reduced to an insignificant minority.

This is also the first study with a detailed discussion of reactions expected to occur during the synthesis of the RGO/Al2O3 nanocomposite. It is widely accepted that the functional groups present on the surface of GO, including carboxylic, hydroxyl, carbonyl, and epoxy groups, can be utilized as nucleating sites to control the size, morphology, and crystallinity of the nanoparticles grown. Our results show that triethyl aluminium undergoes oxidation and hydrolysis at the same time to yield an organic precursor. When H2O and O2 are present deficiently in the reaction only due to the exposure of the reaction mixture to air (like in the R4 process), the mild hydrolysis–oxidation-association process occurs. Because there is an excess of triethyl aluminium in the reaction, association also occurs but without the creation of cage-complexes due to no excess of water.