Preparation and characterization of alkylamine-intercalated graphite oxides

Carbon 45 (2007) 1005–1012 www.elsevier.com/locate/carbon

Preparation and characterization of alkylamine-intercalated graphite oxides Yoshiaki Matsuo *, Tadaaki Miyabe, Tomokazu Fukutsuka, Yosohiro Sugie Department of Materials Science and Chemistry, Graduate School of Engineering, University of Hyogo, 2167 Shosha, Himeji, Hyogo 671-2280, Japan Received 2 June 2006; accepted 19 December 2006 Available online 27 December 2006

Abstract Various n-alkylamine-intercalated graphite oxides were synthesized in the presence of a small amount of hexane and their compositions, chemical bonding and orientation of alkyl chains in them were characterized. The interlayer space of graphite oxide was saturated when about 11 mmol/g of n-hexadecylamine was included in it. Three types of alkylamines in GO were identified in the resulting intercalation compounds, hydrogen-bonded neutral amines, hydrogen-bonded protonated amines and ionically bound protonated amines, based on the infrared and XPS studies. The equilibrium between these species was established. Immersion of GO saturated by n-hexadecylamine in ethanol resulted in the de-intercalation of amines and the amount of residual amines was 3.0 mmol/g, which was rather similar to that of readily exchangeable protons in GO of 3.5 mmol/g. When alkyl chain length became shorter, the resulting intercalation compounds contained less alkylamine due to insufficient hydrophobic interaction between alkyl chains. The X-ray diffraction and polarized-infrared spectroscopic data indicated that alkyl chains of alkylamines took interdigitated monolayer and bilayer orientations for smaller and larger interlayer spacings, respectively.  2006 Elsevier Ltd. All rights reserved.

1. Introduction Recently, the space with nano or sub-nanometer size in solid materials such as that in zeolite and layered materials has been used as reaction media for photochemical reaction, matrices for photofunctional molecules, adsorbent of non-ionic organic chemicals (NOCs) and so on [1–6]. In case of layered materials, various interesting results have been reported. These include the inclusion of ionically bonded organic dyes at high concentrations with controlled aggregation state [2,7], reversible photo-isomerization of diarerethene [8], spiropirane [9], azobenzene derivatives [10], selective photochemical dimerization of ionically bonded organic molecules [11–13] and so on. When the amphiphilic molecules are introduced into the layered materials, the two-dimensional interlayer is filled with the

*

Corresponding author. Tel./fax: +81 79 267 4898. E-mail address: [email protected] (Y. Matsuo).

0008-6223/$ - see front matter  2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.carbon.2006.12.023

hydrophobic part of these molecules, which provides a hydrophobic medium. Accordingly, various neutral molecules were included in the resulting organic interlayers by hydrophobic interaction. Based on these properties, the suppression of the aggregation of organic dyes [14], selective adsorption of non-ionic organic molecules [15–17], photochemical isomerization of azobenzene [18,19], etc. have been reported. The properties of the space such as polarity, size and free volume are important factors to determine if such space is useful or not. These determine how many or what kind of guest molecules can be introduced in the space and the mobility of the introduced molecules. In this context, recently we have found that graphite oxides hydrophobized by alkyltrimethylammonium ions or alkylamines can selectively include various aromatic molecules [20,21] control aggregation state and orientation of organic dyes [22], and photochemical reaction proceeds efficiently in them [23–25]. For the fundamental properties of these host materials, detailed studies of alkyltrimethylammonium ion-intercalated graphite oxides showed that

Y. Matsuo et al. / Carbon 45 (2007) 1005–1012

alkyltrimethylammonium cations are ionically bonded to C–O groups of graphite oxide formed by dissociation of acidic C–OH groups [26,27]. Therefore, the amounts of intercalated ammonium ions depend on the ion exchange capacity of graphite oxide and the size of the ammonium cations. On the other hand, concerning the fundamental properties of alkylamine-intercalated graphite oxides such as chemical bonding between alkylamine and graphite oxide, composition of them, etc., sufficient data have not yet been provided [21,28], though some are reported by Bourlinos et al. for graphite oxide intercalated by alkylamines with shorter alkyl chain lengths [29]. It appeared that higher amounts of alkylamines are introduced into graphite oxide than those observed for alkyltrimethylammonium ions [21], which suggests different interaction between alkylamine and graphite oxide other than ionic bonding. In this study, therefore, we have investigated the properties of alkylamine-intercalated graphite oxides in detail. 2. Experimental Graphite oxide was obtained from natural graphite powder (57– 74 lm) based on Brodie’s method [30]. It was oxidized by potassium chlorate in fuming nitric acid for 3 h at 60 C and the resulting solution was put in excess water. The precipitate was filtered off and then dried at 60 C overnight. This procedure was repeated for 5 times and the composition of C8O4.2H1.4 Æ 0.94 H2O was obtained based on the elemental analysis of carbon and hydrogen. The water content was determined from the weight decrease below 200 C observed by thermogravimetric (TG) analysis. The amount of exchangeable acidic groups in the obtained GO was determined by the back titration of sodium hydroxide solution (0.05 M, 50 ml) of GO (99.5 mg) by 0.05 M hydrochloric acid solution. All the chemicals including potassium chlorate, fuming nitric acid, sodium hydroxide solution, hydrochloric acid solution, n-alkylamines chloroform and cyclohexane were purchased from Nacalai tesque Co. Ltd. and were used as received. Intercalation of n-alkylamines (CnH2n+1NH2; abbreviated as Cn, n = 4, 8, 12, 16) was performed by grinding the mixture of GO (100 mg) and Cn in a pestle in the presence of a small amount of hexane (ca. 1 ml) at room temperature until hexane was evaporated. The resulting samples were kept at 60 C overnight. The nominal Cn/GO molar ratio was changed between 0.7 and 2.4. When the nominal Cn/GO was smaller than 0.7, unreacted GO remained in the sample. The compositions of the obtained C16-intercalated GOs (hereafter (Cn)xGO; x: Cn/GO ratio) were determined from the data of elemental analysis of carbon, nitrogen and hydrogen by combustion method performed at Center of Organic Elemental Microanalysis of Kyoto University. The content of oxygen was calculated by subtracting those of carbon, nitrogen and hydrogen from 100%. The resulting samples were analyzed by X-ray diffraction (Rigaku Rint-2100), FT-IR spectroscopy (Nicolet Avatar-360, KBr method), Xray photoelectron spectroscopy (XPS; Shimadzu ESCA-3400) and TG measurements (Shimadzu TGA-50). TG measurement was performed under air with the temperature increase rate of 5 C/min between room temperature and 800 C. XPS data were recorded after drying the sample under vacuum at room temperature overnight. The binding energies observed were corrected based on those of Au7=2 4f electrons. The powder (C16)0.93GO sample was submersed in a mixture of cyclohexane and chloroform (1:1 by volume, 2 mg/ml) and a homogeneous solution was obtained. This solution was cast on a silicon substrate and thin film (thickness about 1 lm) samples were obtained. The SEM image showed that the layers of the (C16)0.93GO sample were deposited parallel to the substrate. The angle of the C16 chains against the GO layer was determined by polarized IR spectroscopy [31] and X-ray diffraction for

the obtained film samples. The tilt angle relative to the normal on the GO layer planes, c, of the transition moment of the –CH2– stretching vibration of alkylamine against the GO layer was calculated from the dichroic ratio, Ry 0 x 0 , using the following formula for trans alkyl chains:
2 2 2 Ay 2  n22 sin a1  n32 sin a1  1 sin c Ry 0 x0 ¼ ð1Þ ¼ Ax 2  sin2 c where n indicates the refractive index, a1 is the incident angle of the polarized light, and 90 was used for the angle of the optical transition moment of the –CH2– stretching vibration against the GO layer [31].

3. Results and discussion 3.1. Characterization of GO The interlayer spacing of the obtained GO was 0.67 nm, which is typical for that prepared by Brodie’s method and the color was light brown. The pH titration curve of GO dissolved in 0.05 M NaOH aqueous solution toward H+ ion is given in Fig. 1. When 21.6 mmol of H+ was added to the well dispersed colloidal solution of GO, the pH value of the solution reached 7 and inflection point appeared. Therefore, the exchangeable acidic groups in GO was evaluated as 3.5 mmol/g (0.63 mol/GO). This value was comparable to or slightly smaller than those reported by Cassagneau et al. [32] (3.24–6.04 mmol/g), Liu et al. [27] (4.4 mmol/g) and Scholz and Boehm (about 6 mmol/g) [33]. Recent report by Szabo´ et al. [34] pointed out that the cation exchange capacity or acidity of GO determined from titration varies in a wide range, depending on the pH and ionic strength. The exchangeable acidic groups in GO determined in this method, however, was similar to that of the maximum amount of the exchanged hexadecylpyridium cation (4.0 mmol/g) [26]. 3.2. Preparation and composition of (C16)xGO Table 1 summarizes the compositions of (C16)xGO samples calculated from the data of elemental analysis of carbon, hydrogen and nitrogen. Here, the x values were determined as the molar ratios of nitrogen to eight carbons 12 11 10 9

pH

1006

8 7 6 5

0

5

10

15

20

25

H+ added / mmol/g Fig. 1. The pH back titration curve of GO dissolved in 0.1 M NaOH aqueous solution toward H+ ion.

Y. Matsuo et al. / Carbon 45 (2007) 1005–1012

1007

Table 1 Compositions of C16-intercalated graphite oxides prepared with various nominal C16/GO ratios Nominal C16/GO ratioa

C/%

H/%

N/%

O/%

Composition

C16 content/mmol/g

0.7 0.9 1.2 1.5 1.7 1.8 2.1 2.4

69.51 70.41 71.28 73.69 72.98 73.25 73.18 73.40

8.12 8.85 9.45 10.28 10.61 10.72 10.76 11.25

2.89 3.12 3.35 3.71 3.94 4.09 4.21 4.24

19.48 17.62 15.92 12.32 12.48 11.94 11.85 11.11

(C16)0.67C8O4.0H1.4 Æ 0.20C6H14 (C16)0.78C8O3.9H1.4 Æ 0.28C6H14 (C16)0.93C8O3.8H1.4 Æ 0.33C6H14 (C16)1.14C8O3.3H1.4 Æ 0.37C6H14 (C16)1.46C8O4.0H1.4 Æ 0.20C6H14 (C16)1.66C8O4.3H1.4 Æ 0.34C6H14 (C16)1.90C8O4.7H1.4 Æ 0.28C6H14 (C16)1.96C8O4.5H1.4 Æ 0.40C6H14

4.2 4.9 5.9 7.6 9.0 10.0 11.0 11.6

The values in the parentheses are C16/GO ratio based on anhydrous GO.

in GO, assuming that all the nitrogen content was ascribed to the intercalated C16. The compositions of GO in C16intercalated GO samples were calculated from the contents of carbon, hydrogen and oxygen obtained by subtracting those derived from C16. They slightly varied depending on the samples and possessed more hydrogen than that of pristine GO, while oxygen contents were rather similar to that of anhydrous GO. This suggested that water molecules were removed from the layer of GO and instead, hexane used as a solvent remained in the samples. This idea is supported by the reports showing that large amounts of organic compounds such as ethylene gas [35] and pyrene [21] are adsorbed in the interlayer of the intercalation compounds containing long alkyl or perfluoroalkyl chains. Therefore, the composition was determined assuming that all the excess hydrogen came from the hexane remaining in the (C16)xGO samples. The C16 contents in GO were slightly lower than those calculated from the nominal C16/GO ratios and became almost constant at a value of 1.9 for the samples with nominal C16/GO ratios of 2.1 or larger, indicating that the interlayer spacing of GO was saturated by C16 molecules. This saturation amount of C16 in GO (about 1.9 mol/GO or 10.5 mmol/g) was about twice as large as that observed for hexadecyltrimethylammonium ion-intercalated GO (0.9 mol/GO or 5.6 mmol/g) with the same alkyl chain length [26,27]. Further addition of C16 molecules resulted in the (C16)xGO sample containing unreacted white C16 powder. Fig. 2 shows the X-ray diffraction patterns of (C16)xGO samples with various C16

(I):C16 (H):x=1.96 (G):x=1.90 (F):x=1.66 (E):x=1.46 (D):x=1.14 (C):x=0.93 (B):x=0.78 (A):x=0.67

Intensity, a.u.

Ic=5.02 nm Ic=5.02 nm Ic=5.08 nm Ic=5.02 nm (2.76nm) Ic=4.80 nm (3.00nm) Ic=3.03 nm Ic=2.78 nm Ic=2.73 nm

1

3

5

7

9

2 θ / deg. CuK

11

13

contents, together with that of C16. The diffraction peaks due to C16 were observed at 2h = 1.98, 2.24, 4.40, 6.62, 8.80, 10.96 and 13.2, which disappeared after reaction with GO. This indicates that unreacted C16 remaining on the surface of (C16)xGO was negligible as was observed for the reaction of C16 with vanadium xerogel [36]. As reported previously [21], the (C16)xGO samples basically possessed two different interlayer spacings of 2.73–3.03 nm and 4.80–5.08 nm in which alkylamine molecules probably take interdigitated monolayer and bilayer orientations, respectively. The orientation of C16 molecules are discussed in detail later. When the C16 contents were 0.93 or lower, or 1.66 or higher, single phases were obtained, otherwise two phases with different interlayer spacings existed at the same time. 3.3. Chemical bonding of (C16)xGO Fig. 3 shows the IR spectra of (C16)xGO samples with various compositions, together with those of GO and C16. In addition to the absorption peaks due to –CH2– symmetric and asymmetric stretching (2960, 2933, 2918 and 2854 cm1), –CH2– twist (1390 and 1330 cm1), –CH3 (1476 and 723 cm1) and GO (1090, 997 and 846 cm1) vibrations, that at 1654 cm1 was observed for (C16)0.67GO and (C16)0.78GO. The C16 absorption peaks at 1583 and 1180 cm1 were also observed for (C16)1.66GO and (C16)1.87GO. The spectrum of (C16)1.90GO containing a larger amount of C16 became similar to that of C16, except for the absence of sharp

GO (A):x=0.67 (B):x=0.78 (C):x=1.66 (D):x=1.90

Transmittance, a.u.

a

(0.8) (1.0) (1.3) (1.7) (1.8) (2.0) (2.3) (2.6)

C16

15

Fig. 2. X-ray diffraction patterns of (C16)xGO with various C16 contents, together with that of C16. The values in the parentheses are the interlayer spacings of a minor phase.

4000 3600 3200 2800 2400 2000 1600 1200 800

400

wavenumber / cm-1 Fig. 3. IR spectra of (C16)xGO with various C16 contents: (A) x = 0.67, (B) x = 0.78, (C) x = 1.66 and (D) x = 1.90.

1008

Y. Matsuo et al. / Carbon 45 (2007) 1005–1012

peaks at 3340, 3290, 3190 and around 900–1200 cm1. The absorption peaks at 3640 and 1380 cm1derived of hydroxyl groups of acidic character together with that at 1720 cm1due to carboxyl groups [33] at the edge disappeared in the samples containing C16. This indicates that C16 molecules interact with these groups in GO. The peaks at 1654 and 1583 cm1 are derived of amino groups of alkylamines, though the former may be partly superimposed on that of adsorbed water in GO at 1620 cm1. These are assigned to symmetric and asymmetric deformation bands of hydrogen-bonded NH2 [37]. For (C16)0.67GO and (C16)0.78GO, a weak peak at 1525 cm1 was also observed, which can be ascribed to the symmetric or asymmetric deformation band of NHþ 3 [37]. This peak would be hindered by stronger absorption of hydrogen-bonded NH2 for the samples with higher C16 contents. It has been reported that the main oxygen containing functional groups in GO include acidic hydroxyl and ether groups [33]. While the absorption peak at 1090 cm1 due to ether group was almost unchanged along with the increase of C16 molecules in GO, a broad peak centered at 1380 cm1 from hydroxyl group became much weaker. This indicates that acidic hydroxyl groups in GO mainly reacted with C16 molecules and formed the above NHþ 3 and hydrogen-bonded NH2 groups. In addition to this, based on the unchanged absorption peak at 1090 cm1, the nucleophilic attack of amines on the epoxy groups of GO, forming C–N bonding onto the surface of GO layer which was observed by Bourlinos et al. [29] did not occur in our system. Fig. 4 shows the N1s XPS spectra of (C16)xGO samples with various compositions. A broad peak was observed and this was deconvoluted into two peaks at 399.5 and

400.7 eV for (C16)0.67GO and (C16)0.78GO. Ar+ ion sputtering did not result in a change of the spectra, when the XPS measurement was conducted for film sample. These two peaks are assigned to hydrogen-bonded and protonated amines, respectively based on the literature [37,38]. The latter peak, however, appeared at higher binding energies of 401.7 eV in the spectrum of C16-intercalated vanadium xerogel [36], and rather similar to that observed for alkylamines adsorbed on silica from a solution containing the corresponding ammonium salt (400.1 eV) at a lower concentration [37]. The authors suggested that this intermediate binding energy was due to nitrogen atoms in the ammonium groups weakly hydrogen bonded to hydrogen atom of silanol groups. Therefore, we assigned the above peak at 400.7 eV to that from nitrogen atoms in ammonium group weakly hydrogen bonded to the GO layer, though the detailed chemical structures and formation process of them are not clear at this moment. One possible structure based on the literature is shown in Fig. 5B. On the other hand, for (C16)1.66GO and (C16)1.90GO, the peak was deconvoluted into three peaks at 390.0, 400.3 and 401.8 eV. The last peak is apparently caused by NHþ 3 groups (Fig. 5C) and the former two peaks would be due to hydrogen-bonded NH2 (Fig. 5A) and hydrogen-bonded protonated amines (for example Fig. 5B), respectively, though they slightly shifted to lower binding energies. These results indicate that the interaction between C16 and GO is similar to that between silica and alkylamine with long alkyl chains [37]. In the layer of GO, both hydrogen bonded-neutral amines and protonated amines coexist and equilibrium between them appears to be established. When the (C16)xGO sample with excess C16 (C16/ GO = 5.0) was washed with ethanol several times, the

(B):x=0.78

Intensity (A.U.)

Intensity (A.U.)

(A):x=0.67

397

398

399

400

401

402

403

404

397

398

399

Binding energy / eV

(C):x=1.66

401

402

403

404

402

403

404

(D):x=1.90 Intensity (A.U.)

Intensity (A.U.) 397

400

Binding energy / eV

398

399

400

401

Binding energy / eV

402

403

404

397

398

399

400

401

Binding energy / eV

Fig. 4. XPS spectra of (C16)xGO with various C16 contents: (A) x = 0.67, (B) x = 0.78, (C) x = 1.66 and (D) x = 1.90 in N1s region.

Y. Matsuo et al. / Carbon 45 (2007) 1005–1012 (B)

(A)

C16H33

C16 H33

C16 H33

N

N

N+ H

H

O–

O

H O

(A)

H

O–

O

O

(B)

C16H33 N+

H

H

O

(C)

(A)

C16H33

C16 H33

N+

N

H

O–

H O

H O

GO layer

GO layer lower amine content

higher amine content

Fig. 5. Models for the interaction between C16 and GO: (A) hydrogenbonded-neutral C16, (B) hydrogen-bonded-ammonium type C16 and (C) ionically bonded-ammonium type C16. Dotted lines indicate the hydrogen bonding.

resulting sample showed an almost identical N1s XPS spectrum to those obtained for (C16)0.67GO and (C16)0.78GO. In addition to this change, the interlayer spacing greatly decreased from 5.02 to 1.50 nm. Some of the neutral hydrogen-bonded C16 (Fig. 5A) were replaced by ethanol molecules and they were then removed from the layer in the washing or drying process. As the result, the equilibrium shifted toward a decrease of ionically bound C16 (Fig. 5C) and regeneration of hydrogen-bonded protonated C16 (Fig. 5B and increase of hydrogen-bonded neutral C16 (Fig. 5A). The composition of the immersed sample calculated from the elemental analysis of hydrogen; 6.95%, carbon; 69.22% and nitrogen; 2.59% was (C16)0.52GO. The C16/GO ratio of 0.52 (3.3 mmol/g) in this sample was very similar to the amount of acidic groups in GO (3.5 mmol/g-GO or 3.8 mmol/g-anhydrous GO) determined by back titration. 3.4. Thermal properties of (C16)xGO Fig. 6 shows the TG curves of (C16)0.67GO and (C16)1.90GO, together with those of GO and C16. After observing a small weight decrease due to adsorbed water, the weight of the samples started to decrease beginning at 125 and 145 C, respectively, for (C16)0.67GO and (C16)1.90GO. These temperatures where the strong weight decrease began were similar to that observed for C16 (140 C) rather than that for GO (240 C). Therefore, the

1009

weight decrease below these temperatures should be due to the loss or decomposition of C16 molecules in GO. The second weight decrease was observed above 260 C, which would be mainly due to the elimination of oxygen functionalities from GO layers, leaving residual carbon. The weight decrease below 230 C for (C16)0.67GO and (C16)1.90GO had almost the same value of 27%. Assuming that all of this is due to the loss of C16 molecules, 54 and 36% of them were lost from the intercalation compounds, respectively. The remaining C16 molecules were desorbed above 260 C, together with the decomposition of GO layer. The two step desorption was commonly observed for zirconium phosphate intercalated by n-alkylamines with longer alkyl chain lengths and high amine loadings [39]. 3.5. Orientation of C16 molecules in GO The peak positions in the X-ray diffraction pattern of a (C16)0.93GO thin film were very similar to that of the powder, and it contained two phases with interlayer spacings of 2.98 and 4.96 nm. The latter phase was very small and negligible in the analysis shown below. Two types of orientations of alkyl chains in the gallery of GO are possible based on the interlayer spacing of (C16)0.93GO with interlayer spacings of 2.98 nm. One is an interdigitated monolayer orientation and the other is a more inclined bilayer orientation. An all-trans orientation of –CH2– was suggested by the absorption at 2918 cm1 in the IR spectrum for this sample and the alkyl chains are fully extended [40– 42]. Therefore, based on the size of C16 (0.127 · 16 + 0.28 = 2.31 nm) and the thickness of the intercalated species (2.98–0.67 = 2.31 nm), c = 0 and 60 of inclined angles were obtained for monolayer (cos c = 1) and bilayer (cos c = 0.5) orientations, respectively. Fig. 7 shows the dichroic ratio of the absorption of asymmetric vibration mode of methylene –(CH2)n– groups, which decreased as the incident angle increased. This indicates that the tilt angle, c, is less than 54.7 because the dichroic ratio increases with any refractive index, n, if the tilt angle is 1.1

100

w/w0 / %

(A):(C16)

60

0.67 GO

(B):(C16)

1.90 GO

40 20 0

(C):C16 0

100

200

300

400

500

600

700

Temperature / ˚C Fig. 6. TG data of (A) (C16)0.67GO and (B) (C16)1.90GO, together with those of (C) C16 and (D) GO.

Dichroic ratio, Ry’x’

1

(D):GO 80

0.9 0.8 0.7 0.6 0.5

0

10

20

30

40

50

Incident angle, α / degree

60

70

Fig. 7. The dichroic ratio, Ry 0 x 0 as a function of incident angle for (C16)0.93GO thin film on silicon substrate.

Y. Matsuo et al. / Carbon 45 (2007) 1005–1012

higher than this value. Therefore, a bilayer orientation of alkyl chain is excluded, although the refractive index remains unknown. The experimental data agreed well with Eq. (1) using a refractive index of n = 1.5 and tilt angle of c = 0 (interdigitated monolayer) as shown in Fig. 7 (solid line). The refractive index of 1.5 was typical value of organic materials and the use of this value for the C16intercalated vanadium xerogel also provided a satisfactory result as reported in our previous paper [36]. In case of the phase with larger interlayer spacings of 4.80–5.08 nm, the monolayer orientation is excluded because the height of the GO gallery was much larger than that of molecular length of C16. Therefore, C16 molecules should take bilayer orientation with the tilt angles of 27–17 against GO layer. The tilt angles of alkyl chains in various layered materials have been investigated. The favored orientation of the NH2 groups tilts the chains by 34 [43]. However, when van der Waals interaction between alkyl chains becomes dominant, the flexible orientations of NH2 groups do not restrict the tilt angle to distinct values. Therefore, the longer the alkyl chain lengths and the larger the content of alkyl chains become, the smaller the tilt angle of alkyl chains become [43–51]. The higher amine contents in GO are ascribed to the small tilt angles of alkyl chains in the present system. 3.6. Effect of alkyl chain length Table 2 shows the compositions of (Cn)xGO prepared with a nominal Cn/GO ratio of 2.4. The Cn/GO ratios determined from the data of elemental analysis decreased with the decrease of alkyl chain lengths. While they were very close to the nominal Cn/GO ratio for the samples with alkyl chain length longer than 12, it was considerably smaller for those with shorter alkyl chain lengths. This indicates that a large amount of Cn molecules were lost during reaction and the interaction between amine and GO and/or between them was weaker than those for amines with longer alkyl chain lengths. Fig. 8 shows the X-ray diffraction patterns of (Cn)xGO with various carbon numbers in alkyl chains. The diffraction peaks were observed at lower angles when the alkyl chain length became longer. The interlayer spacings of the samples with shorter alkyl chain lengths were very small, 2.48 nm and 0.90 nm for (C8)1.55GO and (C4)0.43GO, respectively. The interlayer spacing of (C4)0.43GO was similar to that reported by Bourlinos et al. [29] which were prepared by the reaction between amine and GO in ethanol/water solution, how-

Ic=5.02 nm

(D):(C16)1.96GO Ic=3.65 nm

Intensity, a.u.

1010

(C):(C12)1.7 GO Ic=2.50 nm (B):(C8)1.55GO Ic=0.90 nm (A):(C4)0.43GO

1

3

5

7

9

2 θ / deg. CuK α

11

13

Fig. 8. X-ray diffraction patterns of (Cn)xGO with various alkyl chain lengths prepared with the nominal Cn/GO ratio of 2.4: (A) (C4)0.43GO, (B) (C8)1.55GO, (C) (C12)1.74GO and (D) (C16)1.96GO.

ever, those of (C12)1.74GO and (C8)1.55GO are much larger. Since the interlayer spacings of the present (Cn)xGO samples decreased, when immersed in ethanol as was observed in Section 3.3, weakly bonded neutral Cn molecules were exchanged by ethanol and ethanol was de-intercalated during the drying process. This resulted in the lower amine contents and accordingly smaller interlayer spacings observed for the ethanol/water system [29]. The increase in the interlayer spacing of 0.23 nm (0.90– 0.67 = 0.23 nm) observed for (C4)0.43GO is not even sufficient to accommodate flat lying amine molecules. When GO was dried at 100 C under vacuum, the interlayer spacing decreased to 0.59 nm because of the loss of water molecules. Based on this value, the increase of interlayer spacing was 0.31 nm and this indicates that the amine molecules lie flat between the layer of GO. However, the C4 content of 0.43 mol/GO is higher than that calculated for flat arrangement of C4, based on the molecular volume of C4 (0.57 · 4 + 0.14 = 0.368 nm2). The area of the C8 formula unit of GO can be calculated as 0.496 · 0.496 · p 3/2 = 0.213 nm2 using the a0 value of 0.248 nm reported for GO [33]. Therefore, the maximum density of cation per C8 in GO with flat lying C4 is calculated to be 0.29 mol/ GO (0.213/2/0.368) [50]. The excess C4 content would be due to the C4 molecules bonded to the acidic functional groups of GO such as carboxylic groups existing at the edge surface. This was apparently observed in the change of IR spectra shown in Section 3.3.

Table 2 Compositions of alkylamine-intercalated graphite oxides with various alkyl chain lengths prepared with the nominal C16/GO ratio of 2.4 Carbon numbers of amines

C/%

H/%

N/%

O/%

Composition

4 8 12 16

58.96 66.46 71.38 73.40

3.94 8.62 10.18 11.25

3.48 5.89 5.01 4.24

33.62 19.03 13.44 11.11

(C4)0.43C8O3.7H1.4 Æ 0.05C6H14 (C8)1.55C8O4.4H1.4 Æ 0.06C6H14 (C12)1.74C8O4.1H1.4 Æ 0.07C6H14 (C4)1.90C8O4.7H1.4 Æ 0.28C6H14

Y. Matsuo et al. / Carbon 45 (2007) 1005–1012

The lower Cn content for the samples with shorter alkyl chain lengths was due to the absence of Cn bonded to GO layers via hydrogen bonding as discussed above. The interaction between Cn and GO layer is almost the same regardless of alkyl chain length, therefore, the hydrophobic interaction between the C4 molecules was insufficient for them to stay in the interlayers of GO. The hydrophobic interaction (van der Waals interaction) between alkyl groups in the layered materials has been systematically investigated by Lagaly and Weiss using layered silicates as host materials [51]. The enthalpy changes during the adsorption of alcohols into dodecylammonium-intercalated layered silicates increased with the increase of alkyl chain length of the alcohols and were in the range of 0.12–0.17 kJ/mol –CH2–. This resulted in the larger expansion of interlayer spacing when the alcohols with longer alkyl chain lengths were co-intercalated. This corresponds well to the weak interaction between GO and ethylenediamine with short alkyl chain length reported previously [52]. 4. Conclusions N-alkylamines with various alkyl chain lengths were successfully intercalated into graphite oxide in the presence of a small amount of hexane. The interlayer space of graphite oxide was saturated with alkyl amine at 11 mmol/g or 1.9 mol/GO in the case of alkyl amines with longer chain lengths. Considerable amounts of hexane used as a solvent seemed to remain in the resulting samples. Three types of alkylamines in GO were identified in the resulting intercalation compounds, hydrogen-bonded neutral amines, hydrogen-bonded protonated amines and ionically bound protonated amines, based on the infrared and XPS studies. The equilibrium between these species seemed to be established. Immersion in ethanol of GO saturated by n-hexadecylamine resulted in the de-intercalation of amines and the amount of residual amines was 3.0 mmol/g, which was rather similar to that of exchangeable proton in GO, 3.5 mmol/g, determined by the back titration. When the alkyl chain length of n-alkylamines decreased, the resulting intercalation compounds contained less alkylamine, even if the nominal amine/GO was high. This was due to the insufficient hydrophobic interaction between the alkylamine molecules between GO layers. The X-ray diffraction and polarized infrared data indicated that alkyl chains of alkylamines took interdigitated monolayer and bilayer orientations for smaller and larger interlayer spacings. The intercalation compounds first decomposed around 200 C, losing alkylamine and then the skeleton of graphite oxide was destroyed at 260 C, together with the further desorption of amines. The high amine contents in GO achieved in this study are very favorable when they are used as matrices of photofunctional molecules, because the stronger spacer effect is expected. It would be easier to control the aggregation state and orientation of photofunctional molecules when they are included in the alkylamine-intercalated GOs with high amine contents.

1011

Acknowledgement The authors wish to thank to Prof. Masayuki Kawaguchi of Osaka Electro-Communication University for his assistance with the XPS measurements and fruitful discussion. References [1] Ogawa M, Kuroda K. Preparation of inorganic–organic nanocomposites through intercalation of organoammonium ions into layered silicate. Bull Chem Soc Jpn 1997;70:2593–618. [2] Shich T, Takagi K. Clay minerals as photochemical reaction fields. J Photochem Photobio C: Photochem Rev 2000;1:113–30. [3] Ramamurthy V. Controlling photochemical reactions via confinement: zeolite. J Photochem Photobio C 2000;1:145–66. [4] Schulz-Ekloff G, Wo¨rle D, Duffel B, Schoonheydt RA. Chromophores in porous silicas and minerals; preparation and optical properties. Micropor Mesopor Mater 2002;51:91–138. [5] Ogawa M. Photo processes in mesoporous silicates prepared by a supramolecular templating approach. J Photochem Photobio C 2002;3:129–46. [6] Boyd SA, Le JF, Mortland MM. Attenuating organic contaminant mobility by soil modification. Nature 1988;333:345–7. [7] Ogawa M, Kuroda K. Photofunctions of intercalation compounds. Chem Rev 1995;95:399–438. [8] Sasai R, Itoh H, Shindaichi I, Shichi T, Takagi K. Photochromism of clay–diarylethene hybrid materials in optically transparent gelatin films. Chem Mater 2001;13:2012–6. [9] Seki T, Ichimura K. Thermal isomerization behaviors of a spiropyran in bilayers immobilized with a linear polymer and a smectitic clay. Macromolecules 1990;23:31–5. [10] Ogawa M. Photocontrol of the basal spacing of azobenzene-magadiite intercalation compound. Adv Mater 2001;13:1107–9. [11] Takagi K, Shich T, Usami H, Sawaki Y. Controlled photocycloaddition of unsaturated carboxylates intercalated in hydrotalcite clay interlayers. J Am Chem Soc 1993;115:4339–44. [12] Takagi K, Usami H, Fukaya H, Sawaki Y. Spatially controlled photocycloaddition of a clay-intercalated stilbazolium cation. J Chem Soc Chem Commun 1989:1174–5. [13] Usami H, Takagai K, Sawaki Y. Controlled photocycloaddition of stilbazolium ions intercalated in saponite clay layers. J Chem Soc Perkin Trans 2 1990:1723–8. [14] Sasai R, Iyi N, Fujita T, Takagi K, Itoh H. Synthesis of rhodamine 6G/cationic surfactant/clay hybrid materials and its luminescent characterization. Chem Lett 2003;32:550–1. [15] Boyd SA, Shaobai S, Le JF, Mortland MM. Pentachlorophenol sorption by organo-clays. Clay Clay Miner 1988;36:125–30. [16] Boyd SA, Mortland MM, Chou CT. Sorption characteristic of organic compounds on hexadecyltrimethylammonium-smectite. Soil Sci Soc Am J 1988;52:652–7. [17] De´ka´ny I, Kru¨ger-Grasser R, Weiss A. Selective liquid sorption properties of hydrophobized graphite oxide. Colloid Polym Sci 1988;276:570–6. [18] Fujita T, Iyi N, Klapyta Z. Preparation of azobenzene–mica complex and its photoresponse to ultra violet irradiation. Mater Res Bull 1998;33:1693–701. [19] Fujita T, Iyi N, Klapyta Z. Optimum conditions for photoresponse of azobenzene-organophilic tetrasilicic mica complexes. Mater Res Bull 2001;36:557–71. [20] Matsuo Y, Hatase K, Sugie Y. Monomer emission from pyrene adsorbed in surfactant-intercalated graphite oxide. Chem Commun 1999:43–4. [21] Matsuo Y, Watanabe K, Fukutsuka T, Sugie Y. Characterization of n-hexadecylalkylamine-intercalated graphite oxides as sorbents. Carbon 2003;41:1545–50.

1012

Y. Matsuo et al. / Carbon 45 (2007) 1005–1012

[22] Matsuo Y, Fukutsuka T, Sugie Y. Preparation and fluorescent properties of Rhodamine B-hexadecylamine-intercalated graphite oxide thin film. Chem Lett 2003;32:1004–5. [23] Matsuo Y, Fukutsuka T, Sugie Y. Photochemical dimerization of acenaphtylene in surfactant-intercalated graphite oxide. Carbon 2002;40:958–61. [24] Matsuo Y, Fukunaga T, Tokura N, Fukutsuka T, Sugie Y. Control of photochemical dimerization of acenaphtylene in graphite oxide intercalated by n-alkylamines with different alkyl chain length. Trans Mat Res Soc Jpn 2004;29:3219–22. [25] Matsuo Y, Fukunaga T, Tokura N, Fukutsuka T, Sugie Y. Photochemical dimerization of acenaphthylene in hydrophobized graphite oxide(II). Trans Mat Res Soc Jpn 2003;28:589–95. [26] Matsuo Y, Niwa T, Sugie Y. Preparation and characterization of cationic surfactant-intercalated graphite oxide. Carbon 1999;3: 897–901. [27] Liu Z, Wang Z, Yang X, Ooi K. Intercalation of organic ammonium ion into layered graphite oxide. Langmuir 2002;18:4926–32. [28] Arago´n F, Cano Ruiz J, MacEwan DMC. b-Type interlamellar sorption complexes. Nature 1959;183:740–1. [29] Bourlinos AB, Gournis D, Petridis D, Szabo´ T, Szeri A, De´ka´ny I. Graphite oxide: chemical reduction to graphite and surface modification with primary aliphatic amines and amino acids. Langmuir 2003;19:6050–5. [30] Brodie MBC. Sur le poids atomique du graphite. Ann Chim Phys 1860;59:466–72. [31] Akutsu H, Kyogoku Y, Nakahara H, Fukuda K. Conformational analysis of phosphatidylethanolamine in multilayers by infrared dichroism. Chem Phys Lipids 1975;15:222–42. [32] Cassagneau T, Guerin F, Fendler JH. Preparation and characterization of ultrathin films layer-by-layer self-assembled from graphite oxide nanoplatlets and polymers. Langmuir 2000;16:7318–24. [33] Scholz W, Boehm HP. Betrachtungen zur struktur des graphitoxids. Z Anorg Allg Chem 1969;369:327–40. [34] Szabo´ T, Tomba´cz E, Ille´s E, De´ka´ny I. Enhanced acidity and pHdependent surface charge characterization of successively oxidized graphite oxide. Carbon 2006;44:537–45. [35] Sumitani M, Tanamura Y, Hiratani T, Ohmachi T, Inoue H. Adsorption of gaseous molecules within polyfluorinated surfactant/ saponite hybrid compound. J Phys Chem Solids 2005;66:1228–33. [36] Matsuo Y, Yamada N, Fukutuska T, Sugie Y. Dispersion of organic dyes in n-hexadecylamine-intercalated vanadium xerogel thin films. Mol Cryst Liq Cryst 2006;452:133–54. [37] Chernyshova IV, Rao KH, Vidyadhar A. Mechanism of long-chain alkylamines on silicates. A spectroscopic study. 1. Quartz. Langmuir 2000;16:8071–4.

[38] Moses PR, Wier LM, Lennox JC, Finklea HO, Lenhard JR, Murray RW. X-ray photoelectron spectroscopy of alkylamine-silanes bound to metal oxide electrodes. Anal Chem 1978;50:576–85. [39] Peeters K, Vansant EF, Carleer R, Mullens J. Thermal decomposition of n-alkylamine a-zirconium phosphate intercalates. Micropor Mater 1995;4:475–87. [40] MacPhail RA, Strauss HL, Snyder RG, Elliger CA. C–H stretching modes and the structure of n-alkyl chains. 2. Long all-trans chains. J Phys Chem 1984;88:334–41. [41] Venkataraman NV, Vasudevan S. Conformation of methylene chains in an intercalated surfactant bilayer. J Phys Chem B 2001;105: 1805–12. [42] Vaia RA, Teukolsky RK, Giannelis EP. Interlayer structure and molecular environment of alkylammonium layered silicates. Chem Mater 1994;6:1017–22. [43] Lagaly G. Crystalline silicic acids and their interface reactions. Adv Colloid Interf Sci 1979;11:105–48. [44] Lagaly G, Weiss Ar. Ein einfaches Verfahren zur Abscha¨tzung der Schichtladung quellungsfa¨higer glimmerartiger Schichtsilicate. Kolloid-Z. Z. Polymere 1967;216/217:356–61. [45] Lagaly G, Weiss Ar. Anordung und Orientierung kationischer Tenside auf ebenen Silicatoberfla¨chen. Teil I: Darstellun der nAlklammoniumderivate von glimmeratigen Shichtsilicaten Kolloid-Z. Z. Polymere 1970;237:266–73. [46] Lagaly G, Weiss Ar. Anordung und Orientierung kationischer Tenside auf ebenen Silicatoberfla¨chen. Teil II Paraffina¨hnliche Strukuturen bei den n-Alkylammonium-Shichtsilicaten mit hoher Schichtladung (Glimmer) Kolloid-Z. Z. Polymere 1970;237: 364–8. [47] Lagaly G, Beneke K, Dietz P, Weiss A. Innerkristallines Reactionsvermo¨gen der Phyllodikieselsa¨ure (H2Si2O5)1. Angew Chem 1974;86:893–4. [48] Weiss Ar, Mehler A, Hofmann U. Zur Kenntnis von organophilem Vermikulit. Z Naturforsch 1956;B11:431–4. [49] Lagaly G. Interaction of alkylamines with different types of layered compounds. Solid State Ionics 1986;22:43–51. [50] Lagaly G. Characterization of clays by organic compounds. Clay Miner 1981;16:1–21. [51] Lagaly G, Weiss A. Zur van-der-Waals wechselwirkung in n-dodeclyammonium-schichtsilicaten. Z Naturforsch 1969;24B: 1057–8. [52] Valerga Jime´nez P, Arufe Martı´nes MI, Martı´n Rodrı´gez A. Studies on ethylenediamine-graphite oxide complexes—structural parameters and interlayer area. Carbon 1985;23:473–9.