**3. The Langmuir‐Blodgett films of graphene oxide derivatives**

As commented previously, due to its unique properties graphene has been suggested to be used in a great number of technological applications. Nevertheless, each application requires a different set of properties. Thus, graphene synthesized by chemical vapor deposition (CVD) or micromechanical exfoliation renders high‐quality sheets suitable for electronic applications; however, these sheets cannot be used for the fabrication of composites or water‐soluble materials, because they do not contain functionalized groups. In these situations, graphene oxides [30] are preferred because they contain reactive oxygen functional groups that can attach small molecules, polymers, or nanoparticles to the graphitic surfaces for potential use in polymer composites [31], gas sensors [32], or photovoltaic cells [33, 34].

Another important issue of the use of graphene in technological applications is related to its implementation in devices. In the particular case of graphene oxide derivatives, conventional deposition techniques such as drop casting [9] or spin coating [10] not only induce aggregation of flakes, as can be seen in **Figure 1**, but also force the sheets to fold and wrinkle, losing its excellent properties [2]. Therefore, to overcome these limitations other deposition techniques such as LB have been recently proposed [16, 35].

Graphene oxide can be considered as an amphiphilic material [36] because it is constituted by two different domains. The hydrophobic one corresponds to π‐conjugated sp<sup>2</sup> carbon while the hydrophilic domain is constituted by O‐groups attached at the basal plane [37]. The existence of two regions allows obtaining stable water‐insoluble monolayers of the material. Accordingly, several properties such as rheological properties, morphology, and stability of GO monolayers have been recently reported [11, 16, 18, 19]. Several works seem to indicate a great influence of the chemical composition on the film properties [19, 38].

Concerning the chemical synthesis, graphene oxide is usually obtained by graphite oxidation [17, 19, 39] or carbon nanofibers [18–20, 40] by means of the Staudenmaier [41] or the Hummers [42] reactions. Then, graphene oxides are often reduced by chemical agents [17, 43, 44] or thermal annealing [45, 46] to restore the graphene structure. However, both the GO reduced by chemical agents, referred as reduced graphene oxide (RGO), and the thermally reduced retain some O‐groups attached to the basal plane of GO. These O‐groups decrease the amazing properties of graphene such as transparency and high electric conductivity.

Despite the interest raised by GO, the knowledge of its chemical structure remains still a challenge. The best‐known graphene oxide structure consists of two different carbon domains constituted by Csp2 corresponding to aromatic groups and Csp3 of alcohol and epoxy groups attached at the basal plane. The carboxylic acid groups are located at the edge of the sheets [37]. However, the main origin of the controversy is the percentage of each group into the flakes. Several are the causes of discrepancies, although the variability of the starting material and the oxidation process seem to be the most important ones [47]. On the other hand, the chemical structure of graphene oxide was recently revisited because it has been proved that the oxidation of carbon‐based materials originates highly oxidized fragment, named as oxidative debris (OD) [47–49]. The oxidized fragments remain strongly adsorbed onto the graphitic sheets due to π‐π staking interactions but can be removed by alkaline washing of graphene oxide. The purified GO contains lower O/C ratio than the non‐purified one, and consequently its chemical structure and solubility properties are quite different [47, 49].

**Figure 2.** Langmuir‐Blodgett (a) and Langmuir‐Schaefer (b) transfer processes.

24 Two-dimensional Materials - Synthesis, Characterization and Potential Applications

When the monolayer is transferred, its structure is often modified; therefore, to construct high‐ quality films, a careful control of experimental parameters, such as stability and homogeneity of the monolayer, subphase properties (composition, pH, presence of electrolytes, and temperature), substrate nature, speed of immersion/emersion of substrate, surface pressure during the deposition process, and the number of transferred monolayers, is required [12, 29].

As commented previously, due to its unique properties graphene has been suggested to be used in a great number of technological applications. Nevertheless, each application requires a different set of properties. Thus, graphene synthesized by chemical vapor deposition (CVD) or micromechanical exfoliation renders high‐quality sheets suitable for electronic applications; however, these sheets cannot be used for the fabrication of composites or water‐soluble materials, because they do not contain functionalized groups. In these situations, graphene oxides [30] are preferred because they contain reactive oxygen functional groups that can attach small molecules, polymers, or nanoparticles to the graphitic surfaces for potential use

Another important issue of the use of graphene in technological applications is related to its implementation in devices. In the particular case of graphene oxide derivatives, conventional deposition techniques such as drop casting [9] or spin coating [10] not only induce aggregation of flakes, as can be seen in **Figure 1**, but also force the sheets to fold and wrinkle, losing its excellent properties [2]. Therefore, to overcome these limitations other deposition techniques

Graphene oxide can be considered as an amphiphilic material [36] because it is constituted by

the hydrophilic domain is constituted by O‐groups attached at the basal plane [37]. The existence of two regions allows obtaining stable water‐insoluble monolayers of the material. Accordingly, several properties such as rheological properties, morphology, and stability of GO monolayers have been recently reported [11, 16, 18, 19]. Several works seem to indicate a

carbon while

two different domains. The hydrophobic one corresponds to π‐conjugated sp<sup>2</sup>

great influence of the chemical composition on the film properties [19, 38].

**3. The Langmuir‐Blodgett films of graphene oxide derivatives**

in polymer composites [31], gas sensors [32], or photovoltaic cells [33, 34].

such as LB have been recently proposed [16, 35].

It is necessary to consider that in nanocomposites built with GO, the second component, polymers, nanoparticles, or small molecules, often interacts with the O‐groups of graphene oxide; therefore, to improve the quality of the composite, it is crucial to have knowledge of the chemical structure of graphene oxide to control interactions between components which have a great influence on the properties of nanocomposites. However, there is no systematic study related to the effect of the oxidation procedure, nature of the starting material, and purification process on the chemical structure and properties of graphene oxides. Recently, we have started the systematic study of the effect of the starting material, reduction protocol, and purification process on the chemical structure of graphene oxides and on the film morphology. With this objective in mind, we have synthesized graphene oxides using graphite, and GANF® nano‐ fibers from the Grupo Antolín Ingenieria (Burgos, Spain) as starting materials. The oxidation procedure was Hummer's reaction modified to obtain more oxidized samples [17, 18–20]. As reducing agents, we used hydrazine, vitamin C, and sodium borohydride. The purification process consisted of alkaline washing of graphene oxide and is previously reported [48, 49].

To quantify the oxidation degree of different materials, X‐ray photoelectron spectroscopy (XPS) was employed. In all samples, the C1s core‐level spectrum is an asymmetric band that can be fitted to three components centered at 284.8, 286.4, and 287.9 eV. These peaks are assigned to C-C bonds of the aromatic network, C**-**O bonds of alcohol or epoxide groups, and COOH groups, respectively [50]. From the area of these peaks, the percentage of the different groups in each sample was calculated. Results obtained for different kinds of graphene oxides are collected in **Figure 3**. Data shown in **Figure 3** were taken from references [17–19].

**Figure 3.** Chemical composition of graphene oxides determined by XPS measurements. Data were taken from referen‐ ces [17–19]. Solid symbols correspond to graphene oxides obtained from graphite and open symbols from GANF® nanofibers. Stars are results of surfactant‐functionalized graphene oxides.

Results in **Figure 3** clearly show differences between the chemical composition of graphene oxides synthesized by the oxidation of graphite (GO) and nanofibers (NGO). Thus, the percentage of Csp2 is slightly higher for NGO than for GO, while the percentage of C**-**O groups at the basal plane is higher for GO than for NGO and the percentage of COOH groups attached to NGO is twice that of GO. This behavior was attributed to the different size of nanoplatelets [18, 19]. In the case of NGO, dynamic light‐scattering measurements (DLS) and the statistical analysis of FESEM images demonstrated that nanoplatelets of NGO are smaller than the GO ones; therefore, since the carboxylic groups are mainly localized at the edge of sheets the smallest sheets contain the highest proportions of COO groups [19]. As far as the influence of the purification procedure on the chemical composition, our results indicated that the per‐ centage of Csp2 increases after the alkaline washing. Moreover, the purification process drives to graphene oxides of similar chemical composition although the chemical structure of non‐ purified graphene oxides is quite different.

Another interesting result is that the percentage of Csp2 of reduced graphene oxide is almost independent on the reducing agent, and the averaged value of 65 ± 2 is lower than the value found for purified graphene oxide, 72 ± 4. This fact was previously reported for graphene oxide reduced by hydrazine [47] and was interpreted as follows: due to the basic nature of hydrazine, it cleans oxidative debris and simultaneously reduces the oxygen groups of graphene oxide; however, nitrogen atoms remain attached to RGO sheets decreasing the percentage of Csp<sup>2</sup> . This C**-**N bond identified by XPS as a peak centered at 400 eV is responsible for the increase of Csp3 percentage. The balance of these processes leads to RGO sheets of intermediate composition between purified PGO and GO [17]. Similar situations were observed for graphene oxides reduced by vitamin C and borohydride, respectively. In these cases, oxygen and boron atoms of the oxidized product of vitamin C and borohydride remain attached to the network decreasing the aromatic degree of graphene derivatives. According to our results, we postulate that alkaline washing must be the preferred procedure to increase the Csp2 percentage on graphene oxide nanoplatelets.

Graphene oxide nanoplatelets are insulators and to increase the electric conductivity chemical reduction has been postulated. However, RGO films prepared by conventional deposition methodologies present low electrical conductivity values. This is probably due to the platelet aggregation induced by dewetting processes. We have explored the Langmuir‐Blodgett methodology to obtain non‐aggregated and ordered reduced graphene oxide films. To prepare the LB film, it is necessary to select the proper surface state, which will be transferred. To identify the surface state of materials at the interface, the compressional modulus, *ε*, has been widely used. The parameter can be calculated from the surface‐pressure isotherm using Eq. (1):

$$
\varepsilon = -A \left( \frac{\delta \Pi}{\delta A} \right)\_{\Gamma, P} \tag{1}
$$

In Eq. (1), *A* represents the surface area and π the surface pressure value. We have recorded the surface pressure isotherms of each material and a representative example is plotted in **Figure 4a**. The compressional elastic modulus value is plotted against the surface pressure in **Figure 4b**.

**Figure 3.** Chemical composition of graphene oxides determined by XPS measurements. Data were taken from referen‐ ces [17–19]. Solid symbols correspond to graphene oxides obtained from graphite and open symbols from GANF®

Results in **Figure 3** clearly show differences between the chemical composition of graphene oxides synthesized by the oxidation of graphite (GO) and nanofibers (NGO). Thus, the

at the basal plane is higher for GO than for NGO and the percentage of COOH groups attached to NGO is twice that of GO. This behavior was attributed to the different size of nanoplatelets [18, 19]. In the case of NGO, dynamic light‐scattering measurements (DLS) and the statistical analysis of FESEM images demonstrated that nanoplatelets of NGO are smaller than the GO ones; therefore, since the carboxylic groups are mainly localized at the edge of sheets the

the purification procedure on the chemical composition, our results indicated that the per‐

to graphene oxides of similar chemical composition although the chemical structure of non‐

Another interesting result is that the percentage of Csp2 of reduced graphene oxide is almost independent on the reducing agent, and the averaged value of 65 ± 2 is lower than the value found for purified graphene oxide, 72 ± 4. This fact was previously reported for graphene oxide reduced by hydrazine [47] and was interpreted as follows: due to the basic nature of hydrazine, it cleans oxidative debris and simultaneously reduces the oxygen groups of graphene oxide; however, nitrogen atoms remain attached to RGO sheets decreasing the percentage of Csp<sup>2</sup>

This C**-**N bond identified by XPS as a peak centered at 400 eV is responsible for the increase

composition between purified PGO and GO [17]. Similar situations were observed for graphene oxides reduced by vitamin C and borohydride, respectively. In these cases, oxygen and boron atoms of the oxidized product of vitamin C and borohydride remain attached to the network decreasing the aromatic degree of graphene derivatives. According to our results,

percentage. The balance of these processes leads to RGO sheets of intermediate

is slightly higher for NGO than for GO, while the percentage of C**-**O groups

increases after the alkaline washing. Moreover, the purification process drives

groups [19]. As far as the influence of

.

nanofibers. Stars are results of surfactant‐functionalized graphene oxides.

26 Two-dimensional Materials - Synthesis, Characterization and Potential Applications

smallest sheets contain the highest proportions of COO-

purified graphene oxides is quite different.

percentage of Csp2

centage of Csp2

of Csp3

The isotherm morphology is similar to that of surfactant molecules and can be interpreted as follows: monolayers of surface pressure close to zero correspond to low values of compres‐

**Figure 4.** (a) Surface pressure and (b) compressional elastic modulus isotherms of graphene oxide reduced by borohy‐ dride at 293 K.

sional modulus and were assigned to surface states in which the nanoplatelets are isolated in a two‐dimensional gas state. When the surface area decreases, the nanoplatelets are pushed closer to each other, resulting in small domains in which ε grows until it reaches a maximum value. This two‐dimensional region is commonly assigned to the liquid‐expanded (LE) state and corresponds to close‐packed sheets. Beyond the compressional elastic modulus maximum, the nanoplatelets form wrinkles, overlaps, and three‐dimensional (3D) structures [16].

In a previous work, the LE state of the GO monolayer [38] has been modeled by Volmer's model adapted to nanoparticles [51]. We have used this model to interpret the isotherms of different nanoplatelets of GO at the LE state. Our results demonstrated strong interactions between carboxylic acids at the edge of sheets through hydrogen bonds [18, 19].

Because we are interested to build GO films of closely packed and nonoverlapped nanoplate‐ lets, we transferred graphene oxide monolayers at the LE state by the LB methodology [18, 19]. Representative atomic force microscopy (AFM) and FESEM images of these films are collected in **Figure 5**.

As can be seen in **Figure 5**, the solid coverage is higher for GO, **Figure 5a**, than for reduced graphene oxides, **Figure 5b**. Low coverage was also reported for purified graphene oxides [18,

**Figure 5.** Representative images of different graphene oxides films: (a) SEM image of graphene oxide obtained by oxi‐ dation of graphite; (b) TEM image of graphene oxide reduced by vitamin C. The inset is a magnification to show the morphology of RGOv nanoplatelets; (c) graphene oxide functionalized with DDPS and reduced by hydrazine. The in‐ set shows a magnification of the AFM image; (d) graphene oxide functionalized with DDPS and reduced by vitamin C. The inset is a TEM image to show details of the film morphology. Reduced graphene oxides were obtained using graphite as starting materials. The surface pressure of the Langmuir monolayer precursor of the LB film was 1 mN m-1.

19] and was attributed to the low percentage of O‐groups attached to sheets of purified oxides [19]. A high percentage of O‐groups favor the contact between silanol groups of the silicon wafer and sheets increasing the adhesion of nanoplatelets to silicon. Since the chemical composition of reduced and purified graphene oxides is almost the same, the low percentage of O‐groups at reduced samples can be responsible for the low coverage observed for RGO films.

sional modulus and were assigned to surface states in which the nanoplatelets are isolated in a two‐dimensional gas state. When the surface area decreases, the nanoplatelets are pushed closer to each other, resulting in small domains in which ε grows until it reaches a maximum value. This two‐dimensional region is commonly assigned to the liquid‐expanded (LE) state and corresponds to close‐packed sheets. Beyond the compressional elastic modulus maximum, the nanoplatelets form wrinkles, overlaps, and three‐dimensional (3D) structures [16].

In a previous work, the LE state of the GO monolayer [38] has been modeled by Volmer's model adapted to nanoparticles [51]. We have used this model to interpret the isotherms of different nanoplatelets of GO at the LE state. Our results demonstrated strong interactions between

Because we are interested to build GO films of closely packed and nonoverlapped nanoplate‐ lets, we transferred graphene oxide monolayers at the LE state by the LB methodology [18, 19]. Representative atomic force microscopy (AFM) and FESEM images of these films are

As can be seen in **Figure 5**, the solid coverage is higher for GO, **Figure 5a**, than for reduced graphene oxides, **Figure 5b**. Low coverage was also reported for purified graphene oxides [18,

**Figure 5.** Representative images of different graphene oxides films: (a) SEM image of graphene oxide obtained by oxi‐ dation of graphite; (b) TEM image of graphene oxide reduced by vitamin C. The inset is a magnification to show the morphology of RGOv nanoplatelets; (c) graphene oxide functionalized with DDPS and reduced by hydrazine. The in‐ set shows a magnification of the AFM image; (d) graphene oxide functionalized with DDPS and reduced by vitamin C. The inset is a TEM image to show details of the film morphology. Reduced graphene oxides were obtained using graphite as starting materials. The surface pressure of the Langmuir monolayer precursor of the LB film was 1 mN m-1.

carboxylic acids at the edge of sheets through hydrogen bonds [18, 19].

28 Two-dimensional Materials - Synthesis, Characterization and Potential Applications

collected in **Figure 5**.

In an attempt to improve the solid coverage, the reduced graphene oxides were functionalized with the zwitterionic surfactant N dodecyl‐N,N‐dimethyl‐3‐ammonio‐1‐propanesulfonate (DDPS). We have proved that the surfactant remains adsorbed onto graphene oxide platelets playing two important roles: as surface active molecule, it favours attractive interactions between the silicon and the reduced graphene oxide and because it is attached at sheets minimizing the restacking of flakes. It is interesting to note that the surfactant is attached to sheets in a non‐covalent way, and consequently the chemical structure of graphene oxide is not significantly altered [52].

The AFM images of functionalized reduced graphene oxide films, **Figure 5c** and **d**, show that the functionalization with the DDPS surfactant increases the solid coverage; however, it is lower than that for graphene oxide, **Figure 5a**. The AFM images of RGOhS, **Figure 5c**, also show the formation of the chained sheets suggesting lateral attractive interactions between flakes. These attractive interactions can be likely induced by the surfactant molecules attached to the sheets [17].

We have great interest to study the effect of GO chemical composition on the electrical conductivity of GO films. However, in the case of reduced graphene oxide the electrical conductivity value is too small to detect significant differences; therefore, we employed an alternative method widely used by other authors. The method consists of measuring the conductivity of paper‐like graphene oxide films [53]. To analyze the electrical conductivity dependence with the chemical composition, we have plotted the electrical conductivity against the Csp<sup>2</sup> and C**-**O group percentages shown in **Figure 6a** and **b**, respectively.

**Figure 6.** Variation of the electrical conductivity values of paper‐like graphene oxide films with the Csp2 (a) and (b) C**-**O group percentages, respectively. Data were taken from Reference [16].

Results in **Figure 6a** show that the electrical conductivity increases as the Csp2 percentage. Moreover, the highest conductivity value is obtained for graphene oxide functionalized with the zwitterionic surfactant. In addition, samples with the lowest percentage of C**-**O and COOH groups, see **Figure 3**, correspond with reduced graphene oxides functionalized with the surfactant DDPS. All these facts suggest that the surfactant molecules can eliminate high amount of O‐groups of samples increasing the electrical conductivity of flakes as can be seen in **Figure 6b**.

On summarizing, the LB technique can be presented as a good methodology of building graphene oxide films because it renders high‐coverage and ordered films. On the other hand, the conductivity of our surfactant‐functionalized RGO samples is higher than the values found in the literature for paper‐like films of reduced graphene oxide [5, 54] functionalized with ionic surfactants, although more efforts must be done to improve the solid coverage and to increase the electrical conductivity values of graphene oxide films.
