*3.2. Modification of rGO-Am with Oligomers of Glutamic Acid*

Modification of rGO-Am was carried out using a "grafting from" technique. One of the common methods for the synthesis of oligomers and polymers of amino acids is the ring-opening polymerization of α-amino acid N-carboxyanhydrides (NCAs) catalyzed with the primary amines [46]. In our case, the primary amino groups of rGO-Am played a role of initiating groups for polymerization of γ-protected glutamic acid NCA to prepare the rGO-Am grafted with oligomers of glutamic acid (oligo(Glu)). The scheme of rGO-Am modification with oligo(Glu) is displayed in Figure 2. At first step, the ring-opening polymerization of Glu(OBzl) NCA was carried out to modify rGO-Am. Then, the Bzl protection was removed from γ-carboxylic groups at acidic conditions. The successful grafting of rGO-Am with oligo(Glu) was testified by FTIR spectroscopy and XPS. *Polymers* **2021**, *13*, x FOR PEER REVIEW 8 of 18

**Figure 2***.* Scheme of rGO-Am modification with oligomers of glutamic acid via ring-opening polymerization of 314 Glu(OBzl) NCA as monomer. 315 **Figure 2.** Scheme of rGO-Am modification with oligomers of glutamic acid via ring-opening polymerization of Glu(OBzl) NCA as monomer.

In comparison with neat rGO-Am and protected and unprotected poly(glutamic ac- 316 id) used as standards, an increase of several characteristic bands corresponding to the 317 groups presented in PGlu was detected in the FTIR spectrum of rGO-Am-oligo(Glu). In 318 particular, an increase in the intensity for the characteristic bands at 1165, 1459, 1561 and 319 1744 cm−1 corresponding to C–N stretching, CH2 bending, N–C=O stretching (amid II) 320 In comparison with neat rGO-Am and protected and unprotected poly(glutamic acid) used as standards, an increase of several characteristic bands corresponding to the groups presented in PGlu was detected in the FTIR spectrum of rGO-Am-oligo(Glu). In particular, an increase in the intensity for the characteristic bands at 1165, 1459, 1561 and 1744 cm−<sup>1</sup> corresponding to C–N stretching, CH<sup>2</sup> bending, N–C=O stretching (amid II)

and С=О stretching vibrations, respectively, was detected after the grafting of oligo(Glu) 321 from rGO-Am (Figure 3). 322

The success of oligo(Glu) grafting from the rGO-Am was also revealed by the corre- 326 sponding changes in the X-ray photoelectron spectra presented in Figure 4a. In the case 327 of both rGO-Am-oligo(Glu(OBzl)) and rGO-Am-oligo(Glu), the rise of the intensity of 328 the O 1s peak, as well as the N 1s peak, was observed. The quantitative analysis of the 329 acquired survey X-ray photoelectron spectra showed that after the ring-opening 330 polymerization of Glu(OBzl) NCA from rGO-Am, the concentration of the nitrogen in- 331 creased from 3.02 at.% to ca. 3.52 and 3.47 at.% for protected and unprotected forms, re- 332

313

323

**Figure 3.** FTIR spectra of neat and modified rGO-Am as well as protected and unprotected poly(glutamic acid) as the 324 standards. 325

and C=O stretching vibrations, respectively, was detected after the grafting of oligo(Glu) from rGO-Am (Figure 3). 1744 cm−1 corresponding to C–N stretching, CH2 bending, N–C=O stretching (amid II) 320 and С=О stretching vibrations, respectively, was detected after the grafting of oligo(Glu) 321 from rGO-Am (Figure 3). 322

**Figure 2***.* Scheme of rGO-Am modification with oligomers of glutamic acid via ring-opening polymerization of 314 Glu(OBzl) NCA as monomer. 315

> In comparison with neat rGO-Am and protected and unprotected poly(glutamic ac- 316 id) used as standards, an increase of several characteristic bands corresponding to the 317 groups presented in PGlu was detected in the FTIR spectrum of rGO-Am-oligo(Glu). In 318 particular, an increase in the intensity for the characteristic bands at 1165, 1459, 1561 and 319

313

323

369

**C=O COOH C–N C/O** 

**Ratio** 

*Polymers* **2021**, *13*, x FOR PEER REVIEW 8 of 18

**Figure 3.** FTIR spectra of neat and modified rGO-Am as well as protected and unprotected poly(glutamic acid) as the 324 standards. 325 **Figure 3.** FTIR spectra of neat and modified rGO-Am as well as protected and unprotected poly(glutamic acid) as the standards.

The success of oligo(Glu) grafting from the rGO-Am was also revealed by the corre- 326 sponding changes in the X-ray photoelectron spectra presented in Figure 4a. In the case 327 of both rGO-Am-oligo(Glu(OBzl)) and rGO-Am-oligo(Glu), the rise of the intensity of 328 the O 1s peak, as well as the N 1s peak, was observed. The quantitative analysis of the 329 acquired survey X-ray photoelectron spectra showed that after the ring-opening 330 polymerization of Glu(OBzl) NCA from rGO-Am, the concentration of the nitrogen in- 331 creased from 3.02 at.% to ca. 3.52 and 3.47 at.% for protected and unprotected forms, re- 332 The success of oligo(Glu) grafting from the rGO-Am was also revealed by the corresponding changes in the X-ray photoelectron spectra presented in Figure 4a. In the case of both rGO-Am-oligo(Glu(OBzl)) and rGO-Am-oligo(Glu), the rise of the intensity of the O 1s peak, as well as the N 1s peak, was observed. The quantitative analysis of the acquired survey X-ray photoelectron spectra showed that after the ring-opening polymerization of Glu(OBzl) NCA from rGO-Am, the concentration of the nitrogen increased from 3.02 at.% to ca. 3.52 and 3.47 at.% for protected and unprotected forms, respectively. Given these differences in the nitrogen content and the known chemical structure of a oligo(glutamic acid) (Supplementary Materials, Section S3), the number of oligo(Glu) branches was calculated and estimated to be ca. 2.2 <sup>×</sup> <sup>10</sup><sup>9</sup> per 1 g of the obtained material (Supplementary Materials, Section S4). The rate of the functionalization of amines were estimated to be about 4–8 branches and 6–12% of the total number of amines, respectively. It means that the average length of the oligo(Glu) chains is in the range from 4 to 8 monomer units. *Polymers* **2021**, *13*, x FOR PEER REVIEW 10 of 18

**Figure 4.** XPS characterization of the rGO-Am grafted with oligo(Glu) (top row) and oligo(Glu(OBzl)) (bottom row): 370 (**a**) survey spectra, (**b**) high-resolution C 1s spectra and (**c**) high-resolution O 1s spectra. 371 **Figure 4.** XPS characterization of the rGO-Am grafted with oligo(Glu) (top row) and oligo(Glu(OBzl)) (bottom row): (**a**) survey spectra, (**b**) high-resolution C 1s spectra and (**c**) high-resolution O 1s spectra.

rGO-Am-oligo(Glu) <0.1 78.79 6.77 6.48 1.65 2.84 3.47 7.38

Binding Energy (eV) 283.7 284.6 285.1 286.8 / 286.4 288.2 289.0 285.9

**C-O-C/C-OH(p)** 

1.93 69.55 12.12 8.01 3.39 1.48 3.52 6.87

*3.3. Characterization of Modified rGO-Am* 374 It is known that chemical modifications of materials can change their characteristics 375 and properties [48–50]. In order to characterize the modified rGO-Am, such physico- 376 chemical methods as DLS and ELS, TGA and DTG were applied for the obtained materi- 377 als. 378 According to the DLS and ELS analyses, the aqueous dispersion of neat rGO-Am 379 (pH 7.4) was characterized by the presence of submicron negatively charged particles 380 with a fairly broad size distribution (Table 3). Modification of rGO-Am with protected 381 and unprotected oligomers was accompanied by a decrease in the hydrodynamic diame- 382 ter of graphene particles due to a reduction of their aggregation. In turn, the modifica- 383 tion of rGO-Am with oligo(Glu(OBzl)) was followed by an increase in the surface elec- 384 trokinetic potential, while, as expected, the deprotection diminished this parameter. 385

**Table 3.** Characteristics of the neat and modified rGO-Am particles in aqueous solution (pH 7.4) 386 determined by dynamic and electrophoretic light scattering. 387 **Sample** *DH* **(nm) PDI ζ-potential (mV)**  rGO-Am 465 ± 71 0.56 −38.4 ± 7.1

rGO-Am-oligo(Glu(OBzl)) 369 ± 66 0.42 −32.1 ± 5.6 rGO-Am-oligo(Glu) 302 ± 49 0.52 −40.0 ± 5.1

**Component C–V C=C C–C C-OH and** 

rGO-Amoligo(Glu(OBzl))

The introduction of oligo(Glu) is further signified by the changes of the C 1s spectra displayed in Figure 4b. As seen, after the synthesis of oligo(Glu) the peaks related to carboxyls and carbonyls centered at *hv* = 289.1 eV and *hv* = 288.2 eV, respectively, increased. The appearance of the C–OH peak at *hv* = 286.2 eV related to phenol groups can be also noted, most probably due to the modification of the graphene layer concurrently with the oligo(Glu) growth. Furthermore, the intensity of the C–C peak at *hv* = 285.1 eV increased owing to the presence of nonconjugated C–C bonds in the oligo(Glu) backbone. As one might expect, the most prominent rise of the C–C peak was observed in the case of rGO-Am-oligo(Glu(OBzl)), where the relative content of the carbon in the corresponding state was up to 12.12 at.% (Table 2) [43]. In turn, the relative concentration of the nonconjugated C–C bonds detected for rGO-Am-oligo(Glu) was almost two times lower, i.e., 6.77 at.%. Combined with the redistribution of the relative content of the carboxyls and carbonyls from 1.48 and 3.39 at.% in rGO-Am-oligo(Glu(OBzl)) to 2.84 and 1.65 at.% in rGO-Amoligo(Glu), respectively, this evolution of the C 1s spectra evidences that, in the latter sample, the protective groups were successfully eliminated from at least 45% of the carboxyl groups in the grafted oligo(Glu). This is also illustrated by the comparative analysis of the O 1s Xray photoelectron spectra displayed in Figure 4c. In both spectra, three peaks are discerned positioned at *hv* = 531.1 eV, *hv* = 532.5 eV and *hv* = 534.2 eV and attributed to double-bonded oxygen, O–H bonded oxygen, and oxygen in the water molecules, respectively [42,43]. As seen, in the case of rGO-Am-oligo(Glu) the peak of the O–H bonded oxygen related to hydroxyl functionality in the deprotected carboxyl group rises with the increase of its relative concentration from 22.39% to 40.48%. Such changes are commonly observed in the case of moving from carbonylated to carboxylated nanocarbon structures [43,47]. Note that the relative content of the adsorbed water also rises from 3.19% in rGO-Amoligo(Glu(OBzl)) to 10.83% in rGO-Am-oligo(Glu) upon the elimination of the protecting Bzl groups and, thus, a higher concentration of hydrophilic carboxyl groups. As a net result, the XPS data implies both the successful grafting of oligo(Glu) from rGO-Am and the following deprotection of up to 45% of the Bzl groups from the carboxyls presented in oligo(Glu).

**Table 2.** Composition of functional groups (at.%) and C/O ratio estimated from the analysis of the deconvoluted C 1s X-ray photoelectron spectra.


#### *3.3. Characterization of Modified rGO-Am*

It is known that chemical modifications of materials can change their characteristics and properties [48–50]. In order to characterize the modified rGO-Am, such physicochemical methods as DLS and ELS, TGA and DTG were applied for the obtained materials.

According to the DLS and ELS analyses, the aqueous dispersion of neat rGO-Am (pH 7.4) was characterized by the presence of submicron negatively charged particles with a fairly broad size distribution (Table 3). Modification of rGO-Am with protected and unprotected oligomers was accompanied by a decrease in the hydrodynamic diameter of graphene particles due to a reduction of their aggregation. In turn, the modification of rGO-Am with oligo(Glu(OBzl)) was followed by an increase in the surface electrokinetic potential, while, as expected, the deprotection diminished this parameter.


**Table 3.** Characteristics of the neat and modified rGO-Am particles in aqueous solution (pH 7.4) determined by dynamic and electrophoretic light scattering.

Neat and modified rGO-Am as well as PGlu and PGlu(OBzl) were analyzed by TGA. The obtained TGA and DTG curves are presented in Figure 5. Before the beginning of the region where the destruction of the carbon base of rGO-Am starts (410–420 ◦C), the mass of the neat rGO-Am consistently falls, starting from 120 to 130 ◦C (Figure 5a). When the intensive thermal destruction process starts (after 400 ◦C), the sample has already lost ~18% of the initial mass. This may probably be due to the removal of the residual oxygenic functional groups and solvent retained between the layers of rGO-Am at the early stages of sample heating. In the case of PGlu(OBzl), the main thermal degradation proceeds in two stages: the sample loses ~59% and 29% of its mass in the regions of 240–360 and 420–570 ◦C, respectively. There is also a low-temperature mass loss in the region of 160–240 ◦C, where a loss of ~4% of mass is noticeable. These three processes are clearly visible on the differential curve of the sample mass change (Figure 5b). The process of thermal destruction of a PGlu sample proceeds in a significantly different way. In this case, three areas of significant mass drop are clearly visible on the TGA and DTG curves: the sample loses ~38% of its mass in the region of 190–320 ◦C; the mass falls by ~18% in the region of 330–400 ◦C; and the sample loses the last ~35% of the mass in the region of 430–560 ◦C. As can be seen when comparing the differential curves for PGlu(OBzl) and PGlu, only the high-temperature mass loss areas (last peaks on the DTG curve, Figure 5b) coincide for the two compared samples. In turn, rGO-Am-oligo(Glu(OBzl)) and rGO-Am-oligo(Glu) were characterized with a more intensive destruction process in the low-temperature region of 200–360 ◦C (the maximum intensity of this process is 290 ◦C) in comparison with neat rGO-Am and PGlu/PGlu(OBzl). Since there are no intense mass losses on the TGA and DTG curves of unmodified rGO-Am in this temperature range, it can be stated that the process observed for the modified samples is the cleavage of oligo(Glu)/oligo(Glu(OBzl)) from rGO-Am and their transition to the gas phase, apparently, with the simultaneous destruction of their oligomer chains. The cleavage of the oligomer fragments from rGO-Am is the most likely due to the identity of the thermodestruction process (the coincidence of temperature regions) for samples containing protected and unprotected oligomer chains. The mass losses for both samples were also very close: ~17% for rGO-Am-oligo(Glu) and 14% for rGO-Am-oligo(Glu(OBzl)). Thus, the mass losses in the region of 200–360 ◦C may be considered as estimated values of the rGO-Am surface grafting with oligomers of glutamic acid: 14–17%.

The analysis of the morphology of the neat and modified rGO-Am platelets did not reveal any obvious surface changes (Figure 6). Both materials demonstrate the corrugated structure with a high number of wrinkles and folds of several µm in scale. This results in the reduction of the π–π\* interlayer stacking between the layers of rGO-Am, resulting in a highly developed surface of the material and irregular porous network structure. This is in contrast to lamellar GO and rGO platelets, having smoothed surface with a low extent of wrinkling and folding. The corrugation of rGO-Am is an intrinsic characteristic of this material related to both the distortion of the graphene layer at nanoscale due to introduced amines and the interaction of amines with the retained oxygenic groups, resulting in folding at a microscale [34]. Notably, it does not depend on the type of solvent used for the deposition varied from the polar ones (isopropyl alcohol) to nonpolar solvents (trichloromethane and tetrachloromethane). This fact is also supported by the absence of alterations of the rGO-Am morphology after its grafting with both PGlu and PGlu(OBzl), modifying its wetting characteristics.

Neat and modified rGO-Am as well as PGlu and PGlu(OBzl) were analyzed by 388 TGA. The obtained TGA and DTG curves are presented in Figure 5. Before the begin- 389 ning of the region where the destruction of the carbon base of rGO-Am starts (410–420 390 °C), the mass of the neat rGO-Am consistently falls, starting from 120 to 130 °C (Figure 391 5a). When the intensive thermal destruction process starts (after 400 °C), the sample has 392 already lost ∼18% of the initial mass. This may probably be due to the removal of the re- 393 sidual oxygenic functional groups and solvent retained between the layers of rGO-Am at 394 the early stages of sample heating. In the case of PGlu(OBzl), the main thermal degrada- 395 tion proceeds in two stages: the sample loses ~59% and 29% of its mass in the regions of 396 240–360 and 420–570 °C, respectively. There is also a low-temperature mass loss in the 397 region of 160–240 °C, where a loss of ~4% of mass is noticeable. These three processes are 398 clearly visible on the differential curve of the sample mass change (Figure 5b). The pro- 399 cess of thermal destruction of a PGlu sample proceeds in a significantly different way. In 400 this case, three areas of significant mass drop are clearly visible on the TGA and DTG 401 curves: the sample loses ~38% of its mass in the region of 190–320 °C; the mass falls by 402 ~18% in the region of 330–400 °C; and the sample loses the last ~35% of the mass in the 403 region of 430–560 °C. As can be seen when comparing the differential curves for 404 PGlu(OBzl) and PGlu, only the high-temperature mass loss areas (last peaks on the DTG 405 curve, Figure 5b) coincide for the two compared samples. In turn, rGO-Am- 406 oligo(Glu(OBzl)) and rGO-Am-oligo(Glu) were characterized with a more intensive de- 407 struction process in the low-temperature region of 200–360 °C (the maximum intensity 408 of this process is 290 °C) in comparison with neat rGO-Am and PGlu/PGlu(OBzl). Since 409 there are no intense mass losses on the TGA and DTG curves of unmodified rGO-Am in 410 this temperature range, it can be stated that the process observed for the modified sam- 411 ples is the cleavage of oligo(Glu)/oligo(Glu(OBzl)) from rGO-Am and their transition to 412 the gas phase, apparently, with the simultaneous destruction of their oligomer chains. 413 The cleavage of the oligomer fragments from rGO-Am is the most likely due to the iden- 414 tity of the thermodestruction process (the coincidence of temperature regions) for sam- 415 ples containing protected and unprotected oligomer chains. The mass losses for both 416 samples were also very close: ∼17% for rGO-Am-oligo(Glu) and 14% for rGO-Am- 417 oligo(Glu(OBzl)). Thus, the mass losses in the region of 200–360 °C may be considered as 418 estimated values of the rGO-Am surface grafting with oligomers of glutamic acid: 14– 419 17%. 420

**Figure 5***.* TGA **(a)** and DTG **(b)** curves for the neat and modified rGO-Am as well as protected and unprotected PGlu as 422 standards. 423 **Figure 5.** TGA (**a**) and DTG (**b**) curves for the neat and modified rGO-Am as well as protected and unprotected PGlu as standards. the absence of alterations of the rGO-Am morphology after its grafting with both PGlu 436 and PGlu(OBzl), modifying its wetting characteristics. 437

**Figure 6***.* SEM images of rGO-Am platelets on the silicon support (×5000): (**a**) neat rGO-Am, (**b**) rGO-Am- 439 oligo(PGlu(OBzl)) and (**c**) rGO-Am-oligo(PGlu). 440 **Figure 6.** SEM images of rGO-Am platelets on the silicon support (×5000): (**a**) neat rGO-Am, (**b**) rGO-Am-oligo(Glu(OBzl)) and (**c**) rGO-Am-oligo(Glu).

#### *3.4. Manufacturing and Characterization of the Composite Films* 441 *3.4. Manufacturing and Characterization of the Composite Films*

In order to prepare the composite materials with rGO-Am and its derivatives, PCL 442 synthesized by ring-opening polymerization of ε-caprolactone was utilized as a polymer 443 matrix in this work. The used PCL had the following characteristics: *M*w = 89,000 and Đ = 444 1.7 (according to SEC), and η = 0.91 dL/g (according to viscosimetric analysis in CHCl3). 445 The films of pure PCL and its composites with neat and modified rGO-Am were 446 In order to prepare the composite materials with rGO-Am and its derivatives, PCL synthesized by ring-opening polymerization of ε-caprolactone was utilized as a polymer matrix in this work. The used PCL had the following characteristics: *M*<sup>w</sup> = 89,000 and Ð = 1.7 (according to SEC), and η = 0.91 dL/g (according to viscosimetric analysis in CHCl3).

polar solvents (trichloromethane and tetrachloromethane). This fact is also supported by 435

421

438

manufactured by casting a polymer solution in CHCl3 on the cellophane substrate as de- 447 scribed in the experimental part (section 2.6). After evaporation of chloroform at room 448 temperature, the films were additionally dried to a constant mass at 45 °C to remove 449 solvent traces. 450 The PCL-based composite films containing neat and modified rGO-Am were tested 451 The films of pure PCL and its composites with neat and modified rGO-Am were manufactured by casting a polymer solution in CHCl<sup>3</sup> on the cellophane substrate as described in the experimental part (Section 2.6). After evaporation of chloroform at room temperature, the films were additionally dried to a constant mass at 45 ◦C to remove solvent traces.

at room temperature in the uniaxial stretching mode. Pure PCL films were used as a 452 benchmark. During the tests, the following characteristics of the material were deter- 453 mined: the modulus of elasticity (or Young's modulus) (*E*), tensile strengths (*σb*) and 454 elongation at break (*εb*). 455 It is known that the aggregation of graphene and its derivatives due to the strong 456 The PCL-based composite films containing neat and modified rGO-Am were tested at room temperature in the uniaxial stretching mode. Pure PCL films were used as a benchmark. During the tests, the following characteristics of the material were determined: the modulus of elasticity (or Young's modulus) (*E*), tensile strengths (*σ<sup>b</sup>* ) and elongation at break (*ε<sup>b</sup>* ).

π–π interaction prevents the homogeneous distribution of these fillers in the polymer 457 matrix. Recently, it has been shown that grafting of rGO with poly (L-lactic acid) in- 458 creases the compatibility of this filler with a poly(L-lactic acid)-based matrix [36]. In- 459 deed, the introduction of neat rGO-Am into the PCL matrix led to a sharply negative re- 460 sult. In particular, a noticeable decrease in the Young's modulus, strength characteristics 461 and deformation resource of the material was observed (Table 4, Figure 7a). 462 It is known that the aggregation of graphene and its derivatives due to the strong π–π interaction prevents the homogeneous distribution of these fillers in the polymer matrix. Recently, it has been shown that grafting of rGO with poly (L-lactic acid) increases the compatibility of this filler with a poly(L-lactic acid)-based matrix [36]. Indeed, the introduction of neat rGO-Am into the PCL matrix led to a sharply negative result. In particular, a noticeable decrease in the Young's modulus, strength characteristics and deformation resource of the material was observed (Table 4, Figure 7a).

**Table 4.** Mechanical properties of PCL and its composites with neat and modified rGO-Am. 463

**Specimen Filler content (%)** *E* **(MPa)** *σ <sup>b</sup>* **(MPa)** *ε <sup>b</sup>* **(%)** 

1.0 348 ± 32 11.4 ± 0.5 12 ± 1


**Table 4.** Mechanical properties of PCL and its composites with neat and modified rGO-Am.

1.0 444 ± 17 14.6 ± 0.4 23 ± 2

**Figure 7.** Stress–strain curves for PCL and its composites with neat (**a**) and modifed (**b**) rGO-Am. 465 **Figure 7.** Stress–strain curves for PCL and its composites with neat (**a**) and modifed (**b**) rGO-Am.

Moreover, the higher the content of the filler added to the matrix, the worse me- 466 chanical properties were detected. The reason for this negative effect is the aggregation 467 of the filler forming large agglomerates in the films. As a consequence, a pronounced 468 heterogeneity of the material structure and the destruction of samples at the phase 469 boundaries took place. The analysis of films by optical microscopy in transmitted light 470 revealed the evident aggregates of rGO-Am in the PCL matrix at filler contents higher 471 than 0.5 wt.% (Figures 8 and 9). At the same time, the modification of rGO-Am with 472 both Bzl-protected and unprotected oligomers of glutamic acid favored a much more 473 uniform distribution of the filler in the PCL matrix due to enhanced compatibility (Fig- 474 ures 8 and 9; Figure S2). In turn, this fact considerably affected the enhancement of the 475 mechanical properties of the manufactured composite materials (Table 4, Figure 7b). The 476 best mechanical properties were established for the PCL-based composite containing 0.5 477 wt.% of rGO-Am-oligo(Glu) as a filler. A similar trend has been recently observed by 478 Wang et al. for PLA-based films filled with neat GO and GO grafted with PLA [36]. The 479 authors found that composites containing neat GO had poorer mechanical properties 480 than nonfilled PLA. At the same time, the modification of GO with PLA contributed to 481 Moreover, the higher the content of the filler added to the matrix, the worse mechanical properties were detected. The reason for this negative effect is the aggregation of the filler forming large agglomerates in the films. As a consequence, a pronounced heterogeneity of the material structure and the destruction of samples at the phase boundaries took place. The analysis of films by optical microscopy in transmitted light revealed the evident aggregates of rGO-Am in the PCL matrix at filler contents higher than 0.5 wt.% (Figures 8 and 9). At the same time, the modification of rGO-Am with both Bzl-protected and unprotected oligomers of glutamic acid favored a much more uniform distribution of the filler in the PCL matrix due to enhanced compatibility (Figures 8 and 9; Figure S2). In turn, this fact considerably affected the enhancement of the mechanical properties of the manufactured composite materials (Table 4, Figure 7b). The best mechanical properties were established for the PCL-based composite containing 0.5 wt.% of rGO-Am-oligo(Glu) as a filler. A similar trend has been recently observed by Wang et al. for PLA-based films filled with neat GO and GO grafted with PLA [36]. The authors found that composites containing neat GO had poorer mechanical properties than nonfilled PLA. At the same time, the modification of GO with PLA contributed to the enchantment of mechanical properties.

464

#### the enchantment of mechanical properties. 482 *3.5. In Vitro Biological Evaluation*

An in vitro biocompatibility study was carried out with the use of human osteoblastlike cells (MG-63 cell line) for both neat and oligo(Glu)-modified rGO-Am as dispersions and as a part of composite films. Figure 10 illustrates the cell viability in the presence of rGO-Am and rGO-Am-oligo(Glu) taken in the range of concentrations from 4 to 1000 µg/mL. As seen, neat rGO-Am was nontoxic up to the concentration of 500 µg/mL, while rGO-Am-oligo(Glu) was biocompatible in the entire tested concentration range. Thus, the modification of rGO-Am by oligomers of glutamic acid improved the biocompatibility of aminated graphene with living cells.

*Polymers* **2021**, *13*, x FOR PEER REVIEW 14 of 18

*Polymers* **2021**, *13*, x FOR PEER REVIEW 14 of 18

**Figure 8.** Images of pure PCL and composite films obatined by optical microscopy in transmitted light (×4): (**a**) PCL; (**b**) 484 PCL/rGO-Am-0.5 wt.%; (**c**) PCL/rGO-Am-1.0 wt.%; (**d**) PCL/rGO-Am-3.0 wt.%; (**e**) PCL/rGO-Am-oligo(Glu(OBzl))-0.5 485 wt.%; (**f**) PCL/rGO-Am-oligo(Glu(OBzl))-1.0 wt.%; (**g**) PCL/rGO-Am-oligo(Glu)-0.5 wt.%; (**h**) PCL/rGO-Am-oligo(Glu)- 486 1.0 wt.%. 487 **Figure 8.** Images of pure PCL and composite films obatined by optical microscopy in transmitted light (×4): (**a**) PCL; (**b**) PCL/rGO-Am-0.5 wt.%; (**c**) PCL/rGO-Am-1.0 wt.%; (**d**) PCL/rGO-Am-3.0 wt.%; (**e**) PCL/rGO-Am-oligo(Glu(OBzl))-0.5 wt.%; (**f**) PCL/rGO-Am-oligo(Glu(OBzl))-1.0 wt.%; (**g**) PCL/rGO-Am-oligo(Glu)-0.5 wt.%; (**h**) PCL/rGO-Am-oligo(Glu)- 1.0 wt.%. **Figure 8.** Images of pure PCL and composite films obatined by optical microscopy in transmitted light (×4): (**a**) PCL; (**b**) 484 PCL/rGO-Am-0.5 wt.%; (**c**) PCL/rGO-Am-1.0 wt.%; (**d**) PCL/rGO-Am-3.0 wt.%; (**e**) PCL/rGO-Am-oligo(Glu(OBzl))-0.5 485 wt.%; (**f**) PCL/rGO-Am-oligo(Glu(OBzl))-1.0 wt.%; (**g**) PCL/rGO-Am-oligo(Glu)-0.5 wt.%; (**h**) PCL/rGO-Am-oligo(Glu)- 486 1.0 wt.%. 487

**Figure 9***.* Images of pure PCL and composite films obatined by optical microscopy in reflected 489 light (×4): (**a**) PCL; (**b**) PCL/rGO-Am; (**c**) PCL/rGO-Am-oligo(Glu(OBzl)); (**d**) PCL/rGO-Am- 490 oligo(Glu). The content of the filler in all composites was 1 wt.%. 491 **Figure 9.** Images of pure PCL and composite films obatined by optical microscopy in reflected light (×4): (**a**) PCL; (**b**) PCL/rGO-Am; (**c**) PCL/rGO-Am-oligo(Glu(OBzl)); (**d**) PCL/rGO-Am-oligo(Glu). The content of the filler in all composites was 1 wt.%.

483

483

488

488

**Figure 10***.* Cytotoxicity study (MTT-test) with MG-63 cells for the dispersions of neat and modified rGO-Am (**a**) as well 510 as PCL-based composite films with rGO-Am and rGO-Am-oligo(Glu) as fillers (**b**). 511 **Figure 10.** Cytotoxicity study (MTT-test) with MG-63 cells for the dispersions of neat and modified rGO-Am (**a**) as well as PCL-based composite films with rGO-Am and rGO-Am-oligo(Glu) as fillers (**b**).

**4. Conclusions** 512 In present work, an approach to the modification of aminated graphene by the 513 "grafting from" technique based on ring-opening polymerization of N- 514 carboxyanhydride of glutamic acid γ-benzyl ester was proposed. Grafting of glutamic 515 acid oligomers was testified by both FTIR spectroscopy and XPS. According to XPS 516 analysis, the approximate chain length of grafted oligomers was in the range of 4–8 517 monomer units. Modification of aminated graphene with oligo(Glu) contributed to a de- 518 The high biocompatibility for polymers of glutamic acid with cells was earlier demonstrated in several papers [38,39,51]. Taking into account the high biocompatibility of PCL and low content of the filler in composites, the absence of cytotoxicity for composite films could be expected. Indeed, PCL-based films filled with rGO-Am and rGO-Am-oligo(Glu) demonstrated a similar and comparable level of cell adhesion to the nonfilled PCL films for all tested composite specimens. Therefore, the addition of rGO-Am into PCL, as well as the modification of rGO-Am, did not affect the biocompatibility of PCL with cells.

509

*3.5. In Vitro Biological Evaluation* 492 An in vitro biocompatibility study was carried out with the use of human osteo- 493 blast-like cells (MG-63 cell line) for both neat and oligo(Glu)-modified rGO-Am as dis- 494 persions and as a part of composite films. Figure 10 illustrates the cell viability in the 495 presence of rGO-Am and rGO-Am-oligo(Glu) taken in the range of concentrations from 496 4 to 1000 µg/mL. As seen, neat rGO-Am was nontoxic up to the concentration of 500 497 µg/mL, while rGO-Am-oligo(Glu) was biocompatible in the entire tested concentration 498 range. Thus, the modification of rGO-Am by oligomers of glutamic acid improved the 499 biocompatibility of aminated graphene with living cells. 500 The high biocompatibility for polymers of glutamic acid with cells was earlier 501 demonstrated in several papers [38,39,51]. Taking into account the high biocompatibility 502 of PCL and low content of the filler in composites, the absence of cytotoxicity for compo- 503 site films could be expected. Indeed, PCL-based films filled with rGO-Am and rGO-Am- 504 oligo(Glu) demonstrated a similar and comparable level of cell adhesion to the nonfilled 505 PCL films for all tested composite specimens. Therefore, the addition of rGO-Am into 506 PCL, as well as the modification of rGO-Am, did not affect the biocompatibility of PCL 507 with cells. 508

#### crease in the average size of rGO-Am particles in the aqueous dispersion. A less pro- 519 **4. Conclusions**

nounced aggregation of modified aminated graphene was also observed after its distri- 520 bution in the PCL matrix. TGA and DTG analysis of modified aminated graphene re- 521 vealed its slightly lower thermal stability in comparison with that of neat rGO-Am. 522 However, the mechanical properties of PCL-based composites obtained with neat ami- 523 nated graphene considerably decreased along with the filler content increase. At the 524 same time, the better distribution of aminated graphene grafted with glutamic acid oli- 525 gomers improved or preserved the mechanical properties of the composite film. The bi- 526 ocompatibility of PCL/rGO-Am-oligo(Glu) composites with human osteoblast-like cells 527 seems to be very promising for their further use as functional scaffolds for bone tissue 528 regeneration. 529 In present work, an approach to the modification of aminated graphene by the "grafting from" technique based on ring-opening polymerization of N-carboxyanhydride of glutamic acid γ-benzyl ester was proposed. Grafting of glutamic acid oligomers was testified by both FTIR spectroscopy and XPS. According to XPS analysis, the approximate chain length of grafted oligomers was in the range of 4–8 monomer units. Modification of aminated graphene with oligo(Glu) contributed to a decrease in the average size of rGO-Am particles in the aqueous dispersion. A less pronounced aggregation of modified aminated graphene was also observed after its distribution in the PCL matrix. TGA and DTG analysis of modified aminated graphene revealed its slightly lower thermal stability in comparison with that of neat rGO-Am. However, the mechanical properties of PCL-based composites obtained with neat aminated graphene considerably decreased along with the filler content increase. At the same time, the better distribution of aminated graphene grafted with glutamic acid oligomers improved or preserved the mechanical properties of the composite film. The biocompatibility of PCL/rGO-Am-oligo(Glu) composites with human osteoblast-like cells seems to be very promising for their further use as functional scaffolds for bone tissue regeneration.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/ 10.3390/polym13162628/s1, Figure S1: The analysis of the size distribution of rGO-Am. (a) The size distribution histogram derived from the laser diffraction measurements. SEM images of (b) an array and (c) a single flake of rGO-Am; S1. Calculations of the molar ratio and number of functional groups in rGO-Am; S2. Calculations of the oligo(Glu) content after the grafting; Figure S2. SEM images of nonfilled PCL and its composite films (×100).

**Author Contributions:** Conceptualization, E.K.-V.; methodology, M.S., M.R. and E.K.-V.; formal analysis, M.S., O.S., M.R., I.G. and E.K.-V.; investigation, M.S., O.S., I.A., I.G., Y.N., G.A., A.S., B.B. and A.N.; data curation, M.S., M.R. and E.K.-V.; writing—original draft preparation, M.R. and E.K.-V.; writing—review and editing, M.R. and E.K.-V.; visualization, M.R. and E.K.-V.; supervision, M.R. and E.K.-V. All authors have read and agreed to the published version of the manuscript.

**Funding:** The work by M.R., G.A., A.S. and B.B. on the XPS studies was supported by the Russian Foundation for Basic Research (grant no. 20-04-60458). The work by M.S., O.S., I.A., I.G., Y.N. and E.K.-V. was carried out as part of a State Assignment.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data are contained within the article and supplementary materials.

**Acknowledgments:** The XPS and SEM studies of aminated graphene and its derivatives were carried out in the Joint Research Center "Materials science and characterization in advanced technology" (Ioffe Institute, St. Petersburg, Russia).

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

