Journal of Colloid and Interface Science 437 (2015) 156–162
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Acceleration of suspending single-walled carbon nanotubes in BSA aqueous solution induced by amino acid molecules Haruhisa Kato a,⇑, Ayako Nakamura a, Masanori Horie b a National Metrology Institute of Japan (NMIJ), National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba Central 5, Higashi 1-1-1, Tsukuba, Ibaraki 305-8565, Japan b Health Research Institute (HRI), AIST, Japan
a r t i c l e
i n f o
Article history: Received 26 July 2014 Accepted 9 September 2014 Available online 19 September 2014 Keywords: PFG-NMR SWCNT Dynamic light scattering Amino acid Suspending Dispersibility Zeta potential
a b s t r a c t Single-walled carbon nanotube (SWCNT) suspensions in aqueous media were prepared using bovine serum albumin (BSA) and amino acid molecules. It was found that the amino acid molecules clearly decreased the time required for suspending the SWCNTs in BSA aqueous solutions. Dynamic light scattering measurements revealed that the particle sizes of the SWCNTs suspended in aqueous media with and without amino acid molecules were approximately the same and stable for more than one week. The zeta potential values of the BSA molecules in pure water and amino acid aqueous solutions were different, and these values were also reflected in the surface potential of colloidal SWCNT particles in the corresponding aqueous media, thus inducing different dispersibility of SWCNTs in aqueous media. Pulsed field gradient nuclear magnetic resonance measurements showed that the interactions between the SWCNTs and the amino acid molecules are weak and comprise chemical exchange interactions and not bonding interactions. Amino acid molecules play a fascinating role in the preparation of SWCNT suspensions in BSA aqueous media by increasing electrostatic repulsive interactions between SWCNT colloidal particles and consequently enhancing the dispersion ability of the BSA molecules. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction Single-walled carbon nanotubes (SWCNTs) have been widely used in chemical and industrial applications, i.e., organic solar cells, transistors, capacitors, drug delivery, and hydrogen storage [1–12]. To utilize SWCNTs for scientific and technological applications, preparation of SWCNT suspensions is a very important step because the bundling structures of SWCNTs strongly affect the electronic structure [13]. However, carbon nanomaterials such as SWCNTs tend to easily self-aggregate through strong van der Waals interactions and hydrophobic interactions in the aqueous phase [14]. Therefore, one of the general techniques for preparing stable SWCNT aqueous suspensions is to suspend SWCNTs using various surfactants such as Triton X, Tween 80, Pluronic F127, sodium dodecyl sulfate, bovine serum albumin (BSA), carboxyl methyl cellulose, and amino acid based amphiphiles under ultrasonication [15–26]. Although chemical functionalization of the SWCNT surface is an effective procedure for suspending SWCNTs in aqueous media individually, surface modification alters the physical and chemical properties of the SWCNTs. Therefore, the ⇑ Corresponding author. Fax: +81 29 861 4618. E-mail address: [email protected] (H. Kato). http://dx.doi.org/10.1016/j.jcis.2014.09.018 0021-9797/Ó 2014 Elsevier Inc. All rights reserved.
suspending procedure for SWCNTs using various surfactants is a very important method. However, appropriate choice of the surfactant is also very important for the preparation of stable and homogeneous aqueous suspensions of SWCNTs. In particular, BSA is a biocompatible dispersant, as shown by in vitro toxicity assessments [27], and works well as an effective dispersant for various nanocarbon materials [28]. However, the time required to suspend SWCNTs in an aqueous phase under ultrasonication is longer than that required for suspending chemically modified SWCNTs in an aqueous phase, resulting in the need for a more effective suspending procedure. The aim of the present study was to develop a procedure for preparing suspensions of SWCNTs within a shorter time using a general biocompatible surfactant, i.e., BSA molecules. The key point of our novel procedure is that we used not only BSA as the dispersant for SWCNTs in the aqueous phase but also amino acid molecules as accelerating agents to enhance the capability of BSA to suspend SWCNTs in aqueous media. The amino acid molecules used in this study are aspartic acid and glutamic acid because they are commercially available and classified as nonessential amino acid molecules. Herein, we discovered a fascinating role of amino acid molecules in the preparation of SWCNT suspensions, and proposed a novel and practical approach that will allow researchers to
H. Kato et al. / Journal of Colloid and Interface Science 437 (2015) 156–162
obtain easily reproducible SWCNT aqueous suspensions using general amino acid molecules. In this study, the dispersibility of the prepared SWCNT suspensions was investigated through dynamic light scattering (DLS) and electrophoretic mobility measurements. The diffusion of amino acid molecules in aqueous media with/without SWCNTs was directly observed by pulsed field gradient nuclear magnetic resonance (PFG-NMR) measurements. Because the PFG-NMR method has the potential to provide a mathematical and physical method for predicting the diffusion phenomena of much smaller targets in solution than DLS, it has been used to study the diffusion phenomena of various surfactant molecules in CNT aqueous suspensions [29–32]; therefore, we employed this method in this study. Using these three measurement methods, we clarified the accelerating effect induced by amino acid molecules on suspending SWCNTs in BSA aqueous solutions, in our novel procedure for the preparation of SWCNT suspensions. 2. Experimental section 2.1. Materials and preparation of aqueous SWCNT suspensions High-purity SWCNT material was synthesized by a waterassisted chemical vapor deposition growth method [33]. The purity and size distribution were carefully evaluated by X-ray Fluorescence spectrometry, transmission electron microscopy observations, and Thermogravimetric Analysis. We used same SWCNT reported by Futaba et al. and the detail properties of this SWCNT are summarized in Ref. [33]. L-aspartic acid (MW = 133.1 g), L-glutamic acid (MW = 147.13 g), and BSA (MW = 66 kDa) were purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). Stable, homogeneous aqueous suspensions of SWCNTs (0.1 mg/mL) were dispersed in D2O aqueous solutions of BSA, L-aspartic acid, and L-glutamic acid using an ultrasonic bath (140 W, 25 kHz). The parameters for the SWCNT/BSA/L-aspartic acid/L-glutamic acid suspensions are summarized in Table 1. 2.2. DLS measurements A DLS7000 (Otsuka Electronics Co., Ltd., Kyoto Japan) with a goniometer system and 45 mW He–Ne laser at a wavelength of 632.8 nm was used. Multiple tau digital correlation was measured at a minimum sampling time of 0.1 ls. The measurements were performed at a scattering angle of 90° with a quartz sample cell. The measurement temperature was regulated at 25.0 °C ± 0.1 °C. The instrument was stored in a clean booth maintained at a constant room temperature of 23.0 °C ± 0.3 °C and humidity of 40% ± 3%. The same sample positions were used for all bottles of SWCNT dispersions for DLS measurements that were stored for longer than one week. Measurements were repeated at least three times, and the mean values were used. Taking the apparent diffusion coefficients obtained by DLS, the hydrodynamic particle sizes of the SWCNT colloidal nanoparticles were determined using the Stokes–Einstein relation, Table 1 Compositions used for SWCNT dispersion. Sample name
SWCNT (mg/mL)
BSA (mg/ mL)
Aspartic acid (mg/mL)
Glutamic acid (mg/mL)
A B1 B2 B3 C1 C2 C3
0.1 – – 0.1 – – 0.1
3.1 – 3.1 3.1 – 3.1 3.1
– 1.2 1.2 1.2 – – –
– – – – 1.2 1.2 1.2
dl ¼
kB T ; 3pgD
157
ð1Þ
where kB is the Boltzmann constant, T is the absolute temperature, g is the viscosity of the solvent, and dl is the calculated hydrodynamic diameter of the colloidal nanoparticles. 2.3. PFG-NMR measurements PFG-NMR measurements were performed on a 14.1 T spectrometer (UNITY INOVA 600A, Varian, California, USA) equipped with an H-F{X} diffusion probe (DSI-V218, Doty Scientific California, USA) capable of producing magnetic field pulse gradients of approximately 2500 G cm1 in the z-direction. The NMR lock was not used for any of these experiments, and the temperature was set at 298.15 K ± 0.1 K. Temperature calibration of the spectrometer was performed using ethylene glycol as a standard. All experiments were carried out using 5 mm o.d. microcell NMR tubes (BMS-005V, Shigemi Co., Ltd., Tokyo, Japan), and the sample height was 5 mm. The diffusion coefficients were calculated using the Stejskal–Tanner diffusion equation,
lnðI=I0 Þ ¼ Dc2 G2 d2 ðD d=3Þ;
ð2Þ
for data with a correlation coefficient of ln(I/I0) vs. higher than 0.99 for the first decay. The PFG stimulated echo (PFGSTE) sequence [(p/2)–s2–(p/2)–s1–(p/2)–s2] acquisition was used [34]. Rectangular pulsed gradients with a duration of 1 ms were incrementally increased from 0 to 595 G cm1 with 64–256 averaged transients. The p/2 pulse width of 12.90 ls, relaxation delay of 10 s, and acquisition time of 2.0 s were defined in advance. The three diffusion times (D) were 50, 75, and 100 ms. Water was purified by a microporous filtration system (Millipore S.A. Japan, Tokyo, Japan) and used as a standard to assure the linearity of the gradient strength from the gradient amplifier. The value used for the water diffusion coefficient was 2.299 109 m2 s1 [35]. 2.4. Electrophoretic mobility measurements Electrophoretic mobility measurements were performed at 25.0 ± 0.1 °C using a zeta potential analyzer (ELSZ, Otsuka Electronics Co., Ltd.). The zeta potentials were estimated in all cases using the simple Smoluchowski equation. Depending on the magnitude of the zeta potential, the spherical Smoluchowski approximation may overestimate the actual zeta potential by up to 20% [36]. 2.5. Raman spectra measurements Raman spectra measurements were performed using a DXR microscope (Thermo Fisher Scientific Inc., Waltham, MA, USA) and recorded with a 532 nm laser line. The SWCNT spectrum presented two peaks at 1580 cm1 and 1340 cm1, which were assigned to G and D bands, respectively. Using these two peaks, the G:D ratios were calculated. Low-energy Raman features that correspond to scattering from the radial breathing mode (RBM) were also used to verify the diameters of SWCNTs. The calculated values were 1.5–1.8 nm. 3. Results and discussion 3.1. Preparation and characterization of SWCNT aqueous colloidal suspensions with and without amino acids The effects of two different dispersants for preparing SWCNT suspensions in aqueous media were investigated as follows. One dispersant was a typical biocompatible dispersant, BSA, and the other dispersant was a mixture of BSA and amino acids. The aqueous
158
H. Kato et al. / Journal of Colloid and Interface Science 437 (2015) 156–162
suspensions of SWCNTs (0.1 mg/mL) were prepared in 3.1 mg/mL aqueous solutions of BSA and BSA/amino acid mixtures (3.1 mg/mL for BSA and 1.2 mg/mL for amino acid molecules) using an ultrasonic bath (140 W, 25 kHz) for 2 h at 15 min intervals. Fig. 1 shows the time required for the preparation of SWCNT suspensions in both the aqueous media. The SWCNT suspension in the aqueous solution of BSA/amino acid (aspartic acid) mixture homogenized and dispersed more rapidly than that in the BSA aqueous solution, in 45 min, as shown in Fig. 1. After 45 min of ultrasonication time, the mean sizes of colloidal SWCNT particles were not changed and took similar/stable sizes. This phenomenon was observed in aqueous solutions of BSA/amino acid mixtures (3.1 mg/mL for BSA and 0.6–1.2 mg/mL for amino acid), however; the required homogenizing time increased at the lower amino acid concentration than 0.6 mg/mL (see Supporting information). In contrast, when SWCNTs were suspended in an aqueous solution with only BSA, more than 100 min was required for the SWCNT aqueous suspension to become homogenous and take stable colloidal size. We also prepared SWCNT aqueous suspensions using glutamic acid instead of aspartic acid and found a similar phenomenon. This result indicates that the amino acid molecules work as accelerators in the preparation of homogenous aqueous suspensions of SWCNTs. DLS measurements were performed for all the SWCNT aqueous suspensions. Although a simple assumption was used (e.g. size distribution, sphere structure. . .) in the DLS analysis for SWCNT colloidal particles, the monitoring of the apparent diameter of colloidal particles by DLS over a period is useful to examine the stability of prepared SWCNT aqueous suspensions and to compare the sizes among suspensions. The hydrodynamic diameters, expressed as light scattering intensity-averaged diameters, are shown in Fig. 2. We used a reverse Laplace transformation for light scattering data analysis because the raw data agreed with the observed photon correlation function, indicating that the SWCNT samples followed an approximate monomodal size distribution.
BSA + amino acid
As shown in Fig. 2, the values obtained from the DLS measurements for SWCNT/BSA and SWCNT/BSA/amino acid suspensions were unchanged for more than one week after preparation. These results indicated that BSA and BSA/amino acid worked well as dispersants for SWCNTs in aqueous media, and the observed hydrodynamic sizes of the SWCNTs were similar for all suspensions, regardless of whether BSA or BSA/amino acid was used. Amino acid molecules worked as accelerating agents that enhanced the capability of BSA to suspend SWCNTs in aqueous media. The zeta potentials of the SWCNT aqueous suspensions were measured to estimate changes in the surface potentials depending on the different dispersant combinations. The uncertainties in the table were calculated from the repeatability of the observed zeta potentials, which were obtained from at least three separate measurements. The zeta potentials of the SWCNTs in the BSA aqueous solution and in aqueous solutions of BSA/amino acid (aspartic acid and glutamic acid) mixtures were obviously different. As shown in Table 2, the zeta potential of BSA molecules was approximately 18 mV and that of SWCNT colloidal particles in BSA aqueous solution (A) was 22 mV, indicating that the BSA molecules were adsorbed on the SWCNTs. However, the estimated value of the zeta potential of SWCNTs in BSA/amino acid aqueous solutions (B3 and C3) was 30 mV for both kinds of amino acid molecules. The zeta potentials of BSA molecules in both amino acid aqueous solutions (B2 and C2) were measured, and the values are shown in Table 2. Interestingly, the zeta potentials of B2 and C2 were similar to those of B3 and C3. The zeta potentials of BSA molecules in amino acid aqueous solutions increased, and the absolute values became larger upon the addition of amino acid molecules, indicating that the amino acid molecules interacted on the surfaces of BSA molecules. This result showed that the surface-modified BSA molecules work as excellent dispersants of SWCNTs in aqueous phases. The zeta potentials of SWCNTs in aqueous solutions of BSA/amino acid (B3 and C3) reflected those of the BSA molecules whose surface potential was modified by amino acids (B2 and C2). According to
BSA
0 min
45 min
90 min
15 min
60 min
105 min
30 min
75 min
120 min
Fig. 1. Photographs of SWCNT suspensions: left figures show SWCNT/BSA/amino acid (aspartic acid) aqueous suspensions, and right figures show SWCNT/BSA aqueous suspensions. Faster dispersion and homogenization, in 45 min, were observed in the SWCNT/BSA/amino acid aqueous suspension, whereas more than 100 min was required to obtain the homogenous SWCNT/BSA aqueous suspension.
159
H. Kato et al. / Journal of Colloid and Interface Science 437 (2015) 156–162 Table 3 G:D ratios of the SWCNTs in various suspensions.
A B3 C3
Sample name
GD ratio
Standard uncertainty
A B3 C3
4.2 4.6 4.0
0.3 0.6 0.2
In Table 3, the G:D ratios for all SWCNT suspensions were calculated from Raman spectroscopic measurements. We could not find any obvious difference between the values for the three suspensions, indicating that the amino acid molecules accelerated the suspending process and did not chemically react to decompose the SWCNTs in the aqueous phases. Fig. 2. Plots of hydrodynamic diameter of SWCNTs vs. time (d). The diameter did not change during the measurement, indicating no agglomeration for SWCNT/BSA and SWCNT/BSA/amino acid suspensions.
Table 2 Zeta potentials of the SWCNT suspensions. Sample name
Zeta potential (mV)
Standard uncertainty (mV)
A B2 B3 C2 C3 BSA
21.5 34.0 29.6 26.5 30.4 18.2
0.6 2.0 1.0 1.4 1.4 1.4
the DLVO (Derjaguin, Landau, Verwey, and Overbeek) theory [37,38], the dispersibility of a material suspension is based on two different interactions, electrostatic repulsive interactions and van der Waals attractive interactions. Since the sizes of colloidal SWCNT particles both in BSA aqueous solution and BSA/amino acid aqueous solution were same as shown in Fig. 2, it was assumed that the van der Waals attractive interactions of A, B3, and C3 were approximately similar. Herein, we therefore focused on the electrostatic repulsive interaction between colloidal SWCNT particles. The electrostatic repulsive energy is not equal to the zeta potential because the latter is the potential for the slip face where a liquid starts to flow inside the electrical double layer formed on the hydrodynamic surface of a material. However, these two values are strongly related to each other. Thus, the absolute values of the zeta potential were larger in B3 than in A, which indicated that the electric repulsive interaction between SWCNT colloidal particles was caused by the amino acid molecules that interacted with the surfaces of the SWCNTs. According to DLVO theory, the repulsive interaction between nanoparticles from the electrical doublelayer can be described by
VðRÞ ¼ pe0 edW2 ejr :
3.2. Characterization of amino acid molecules in aqueous solution/ SWCNT suspensions by PFG-NMR Fig. 3 shows the 1H spectra of the various aqueous solutions/ suspensions (A, B1, B2, and B3) measured with the PFGSTE pulse sequence at a gradient strength of 0 G cm1 and 50 G cm1. The peaks of the BSA molecules broadened and those of the amino acids were sharp. The sharp methine peak (NH2CH) at 3.9–3.7 ppm was close to the water peak, and the sharp peaks from 3.1 to 2.6 ppm were assigned as the methylene protons (CH2). The spectral peak at 3.9–3.7 ppm was sufficiently strong, as shown in the figure, allowing us to carry out PFG-NMR diffusion measurements using this peak. To determine the appropriate parameters for the pulse sequence, we measured the longitudinal (T1) and transverse (T2)
(a)
(b)
ð3Þ
In the above expressions, e0 is the permittivity of vacuum, e is the dielectric constant of the electrolyte solution, e is the elemental charge, kB is the Boltzmann constant, T is the absolute temperature, q/i is the ion number density in the bulk electrolyte, z is the valence of ion i, and j is the inverse of the Debye length. Assuming that the zeta potentials are equal to the surface potentials of SWCNT colloidal particles in Eq. (3), the electrical repulsive energy of B3 and C3 (the zeta potentials were approximately 30 mV) was therefore approximately 2.3 times larger than that of the SWCNTs in the just BSA aqueous solution (the zeta potentials were approximately 20 mV) due to V(R) / W2 in this case. It was estimated that the increased surface charge of the SWCNTs in the suspension in samples B3 and C3 was a significant factor in decreasing the suspending time of the SWCNTs in the aqueous phase.
Fig. 3. Examples of 1H NMR spectra of surfactant solutions collected using the PFGSTE pulse sequence at gradient strengths of 0 G cm1 and 50 G cm1. The diffusion time was 50 ms. (a) A (black), B1 (red), B2 (green), and B3 (blue) at a gradient strength of 0 G cm1; (b) A (black), B1 (red), B2 (green), and B3 (blue) at a gradient strength of 50 G cm1. The peak at 4.7 ppm was assigned to HDO. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
160
H. Kato et al. / Journal of Colloid and Interface Science 437 (2015) 156–162
Table 4 Longitudinal (T1) and transverse (T2) relaxation times of amino acid molecules at 3.8 ppm. Sample name
T1 (s)
T2 (s)
B1 B2 B3 C1 C2 C3
4.6 4.2 3.3 4.5 4.0 3.1
1.3 1.2 1.0 1.2 1.1 0.8
relaxation times at 3.9–3.7 ppm for the typical peak for both amino acids, as shown in Fig. 3. The observed results are summarized in Table 4. The values of T1 and T2 were similar for both amino acid molecules. For these PFG-NMR measurements, the recycle decays were set at 10 s, and the observed diffusion times were set at 50, 75, and 100 ms. The T1 and T2 values slightly changed between B1 vs. B2 and C1 vs. C2, indicating that the BSA molecules and amino acid molecules interacted weakly with each other, since the relaxation value is strongly influenced by slow molecular motion. A larger reduction of the T1 and T2 values was found in the SWCNT/BSA/amino acid suspensions (B3 and C3) than in the BSA/amino acid solutions (B2 and C2), indicating that the amino acid molecules interacted more strongly with the SWCNT/BSA colloidal particles than the BSA molecules. However, the observed differences in the relaxation times among B1/B2/B3 and C1/C2/C3 were small (10–20% difference), and the amino acid molecules individually diffused into the solutions/suspensions. The plots obtained from the PFGSTE echo signal attenuation for the two amino acid molecules are shown in Fig. 4. The attenuation plots were approximately linear and did not depend on the diffu-
sion times, indicating that the two amino acid molecules completely dissolved in water. The diffusion coefficients of the amino acid molecules calculated using linear regression (solid lines, Fig. 4) were 7.34 ± 0.09 1010 and 6.64 ± 0.05 1010 m2 s1 for samples B1 and C1, respectively. The uncertainties in the diffusion coefficients were calculated according to Ref. [39]. The small difference in the diffusion coefficients was induced by the difference in the length of the methylene unit between the two amino acid molecules. The plots obtained from the PFGSTE echo signal attenuation for the two amino acid molecules in BSA aqueous solutions are shown in Fig. 5. The attenuation plots were approximately linear and did not depend on the diffusion times, indicating that the interaction between amino acid and BSA molecules is weak. The diffusion coefficients of the amino acid molecules calculated by linear regression (solid lines, Fig. 5) were 6.75 ± 0.13 1010 and 6.01 ± 0.07 1010 m2 s1 for samples B2 and C2, respectively. The calculated diffusion coefficients of amino acid molecules for B2 and C2 were slightly lower than the corresponding values for B1 and C1, respectively. In addition, as shown in Table 4, the relaxation times for two amino acid molecules were slightly reduced. These results show that the BSA and amino acid molecules interacted weakly and were not strongly bonded to each other, since there was no attenuation peak that agreed with the diffusion coefficient of 7 1011 m2 s1, which is the value for BSA molecules, in the PFG-NMR measurements. Fig. 6 shows the plots of the PFGSTE echo signal attenuation for the two amino acid molecules in BSA/SWCNT aqueous suspensions. Interestingly, the attenuation plots were still approximately linear but did not show the same dependency on the diffusion times. The values of the diffusion coefficients of the amino acid molecules calculated by linear regression (solid lines, Fig. 6) were
(a)
γ
δ Δ δ/3
(a)
γ
δ Δ δ/3
(b)
(b)
γ
δ Δ δ/3
Fig. 4. PFG-NMR spin-echo signal attenuation plots for amino acid aqueous solutions (a) B1 and (b) C1 at d = 1 ms for diffusion times of D = 50 ms (d), D = 75 ms (j), D = 100 ms (N). The solid line represents the linear regression. The diffusion coefficients calculated from the slope are 7.34 ± 0.09 1010 and 6.64 ± 0.05 1010 m2 s1 for B1 and C1, respectively.
γ
δ Δ δ/3
Fig. 5. PFG-NMR spin-echo signal attenuation plots for amino acid/BSA aqueous solutions (a) B2 and (b) C2 at d = 1 ms for diffusion times of D = 50 ms (d), D = 75 ms (j), D = 100 ms (N). The solid line represents the linear regression. The diffusion coefficients calculated from the slope are 6.75 ± 0.13 1010 and 6.01 ± 0.07 1010 m2 s1 for B2 and C2, respectively.
161
H. Kato et al. / Journal of Colloid and Interface Science 437 (2015) 156–162
(a)
Table 5 Diffusion coefficients of aspartic acid and glutamic acid molecules obtained from PFGNMR. Sample name B1 B2 B3 C1 C2 C3
γ
δ Δ δ/3
(b)
γ
δ Δ δ/3
Fig. 6. PFG-NMR spin-echo signal attenuation plots for amino acid/BSA/SWCNT aqueous suspensions (a) B3 and (b) C3 at d = 1 ms for diffusion times of D = 50 ms (d), D = 75 ms (j), D = 100 ms (N). The solid line represents the linear regression. The diffusion coefficients calculated from the slope are 6.51 ± 0.11 1010 and 5.46 ± 0.11 1010 m2 s1 for B3 and C3, respectively.
6.51 ± 0.11 1010 and 5.46 ± 0.11 1010 m2 s1 for samples B3 and C3, respectively. Although very low diffusion PFGSTE signal attenuation was expected because of the strongly adsorbed amino acid molecules on/in the SWCNT colloidal particles, only slightly slower diffusion of the amino acid molecules was observed as compared to those in water or BSA aqueous solutions. As shown in Table 4, a reduction in the relaxation times for the two amino acid molecules is observed in SWCNT suspensions (B3 and C3). However, the changes are not large, thereby supporting the slight changes in the diffusion coefficients of the amino acid molecules (Fig. 6). These results indicated that the amino acid molecules and colloidal SWCNT particles interacted weakly, for example, through a chemical exchange on the surface of the colloidal particles, and that they do not aggregate/agglomerate in aqueous suspensions. Table 5 summarizes the observed diffusion coefficients for the two amino acid molecules under various conditions. A reduction in the diffusion coefficients of the amino acid molecules in the presence of BSA molecules or SWCNT colloidal particles could be clearly seen; however, this reduction was only 10–20%. It is clear that amino acid molecules reduce the suspending time of SWCNTs in BSA aqueous solutions, as shown in Fig. 1. The interaction between the SWCNT colloidal particles and the amino acid molecules is very weak and probably similar to a chemical exchange. Nevertheless, it was observed that the suspending speed of SWCNTs in BSA aqueous solutions was obviously higher with the amino acid molecules than without the amino acid molecules, indicating that surface potential modification of the BSA molecules by the amino acid molecules is a significant force for effectively suspending SWCNTs. This modification results from the increased electrostatic repulsive interactions between SWCNT
Diffusion coefficient (m2 s1) 10
7.34 10 6.75 1010 6.51 1010 6.64 1010 6.01 1010 5.46 1010
Standard uncertainty (m2 s1) 8.91 1012 1.33 1011 1.08 1011 4.76 1012 6.47 1012 1.09 1011
colloidal particles, as observed from zeta potential measurements. Over this decade, surface chemical modifications of SWCNTs have been performed to effectively suspend SWCNTs in aqueous phases [40–42]. However, in this study, it has been shown that surface potential changes of the surfactants are also an important aspect in preparing stable and homogeneous aqueous suspensions of SWCNTs. We could not observe obvious difference in the acceleration effects between aspartic acid and glutamic acid molecules for suspending SWCNTs in BSA aqueous solution, because of the different lengths of the methylene units and the numbers of –COOH group; however, the observed diffusion coefficients were slightly different because of the difference in the molecular weights, as shown in Table 5. Finally, we believe that enhancing the dispersing capability of the surfactant BSA for effectively suspending carbon nanomaterials in aqueous media by means of the increased electrostatic repulsive interactions between the nanomaterials has significant applications in the industrial field. In addition, for in vivo toxicity assessments of SWCNTs, SWCNT bundle aqueous suspensions with various types of surfactants are typically used. However, according to this study, amino acid molecules in the body can be expected to work as accelerators for re-suspending the SWCNT bundles inside the body. This finding is significant in the consideration of the clearance of not only SWCNTs but also other nanomaterials for in vivo toxicity assessments.
4. Conclusions The acceleration effects of amino acid molecules on the preparation of SWCNT suspensions in BSA aqueous solutions were investigated. The time required for suspending SWCNTs while using an ultrasonic bath with amino acid/BSA aqueous solution was clearly shorter than that for the BSA aqueous solution. The zeta potentials of BSA molecules in pure water and amino acid aqueous solutions were different, approximately 20 mV and 30 mV, respectively. These zeta potentials were similar to the surface potential of SWCNT colloidal particles in the corresponding aqueous media, indicating that modification of the surface potential of BSA by amino acid molecules induces differences in the dispersibility of SWCNTs in aqueous media based on electrostatic repulsive interactions. It was also found that the interactions between the SWCNTs and amino acid molecules are very weak and may be a chemical exchange interaction, and not a binding interaction on/in the SWCNTs. Our studies provide significant findings that are expected to be useful for nanomaterial-related research in industrial and biological fields.
Acknowledgments This work was funded by the New Energy and Industrial Technology Development Organization of Japan (NEDO) Grant ‘‘Evaluating risks associated with manufactured nanomaterials (P06041).’’
162
H. Kato et al. / Journal of Colloid and Interface Science 437 (2015) 156–162
Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcis.2014.09.018. References [1] T. Chen, S. Wang, Z. Yang, Q. Feng, X. Sun, L. Li, Z. Wang, Angew. Chem., Int. Ed. 50 (2011) 1815–1819. [2] A. Bianco, K. Kostarelos, C.D. Partidos, M. Prato, Chem. Commun. 57 (2005) 571–577. [3] H. Yoon, J. Ahn, P. Barone, K. Yum, R. Sharma, A. Boghosian, J.H. Han, M.S. Strano, Angew. Chem., Int. Ed. 50 (2011) 1828–1831. [4] J.A. Fagan, M.L. Becker, J. Chun, E.K. Hobbie, Adv. Mater. 20 (2008) 1609–1613. [5] M.A. Bissett, I. Köper, J.S. Quinton, J.G. Shapter, Phys. Chem. Chem. Phys. 13 (2011) 6059–6064. [6] C. Vijiayakumar, B. Balan, M. Kim, M. Takeuchi, J. Phys. Chem. 115 (2011) 4533–4539. [7] Y. Zhao, L. Hu, J.F. Stoddart, G. Grüner, Adv. Mater. 20 (2008) 1910–1915. [8] X. Chen, A. Kis, A. Zettl, C.R. Bertozzi, Proc. Natl. Acad. Sci. 104 (2007) 8218– 8222. [9] A. Rajan, M.S. Strano, D.A. Heller, T. Hertel, K. Schulten, J. Phys. Chem. B 112 (2008) 6211–6213. [10] A. Mateo-Alonso, C. Ehli, K.H. Chen, D.M. Guldi, M. Prato, J. Phys. Chem. A 111 (2007) 12669–12673. [11] F. Ding, P. Larsson, J.A. Larsson, R. Ahuja, H. Duan, A. Rosen, K. Bolton, Nano Lett. 8 (2008) 463–468. [12] R. Marquis, C. Greco, I. Sadokierska, S. Lebedkin, M.M. Kappes, T. Michel, L. Alvarez, J.L. Sauvajol, S. Meunier, C. Mioskowski, Nano Lett. 8 (2008) 1830– 1835. [13] H. Ozawa, T. Fujigaya, Y. Niidome, N. Hotta, M. Fujiki, N. Nakashima, J. Am. Chem. Soc. 133 (2011) 2651–2657. [14] M.J. O’Connell, S.M. Bachilo, C.B. Huffman, V.C. Moore, M.S. Strano, E.H. Haroz, K.L. Rialon, P.J. Boul, W.H. Noon, C. Kittrell, J. Ma, R.H. Hauge, R.B. Weisman, Science 297 (2002) 563–566. [15] A.L. Alpatova, W. Shan, P. Babica, B.L. Upham, A.R. Rogensues, S.J. Masten, E. Drown, A.K. Mohanty, E.C. Alocilja, V.V. Tarabara, Water Res. 44 (2010) 505– 520. [16] H.N. Yehia, R.K. Draper, C. Mikoryak, R.K. Walker, P. Bajaj, I.H. Musselman, M.C. Daigrepont, G.R. Dieckmann, P. Pantano, J. Nanobiotechnol. 5 (2007) 1–17. [17] P. Wick, P. Manser, L.K. Limbach, U. Dettlaff-Weglikowska, F. Krumeich, S. Roth, W.J. Stark, A. Bruinink, Toxicol. Lett. 168 (2007) 121–131.
[18] M. Chiaretti, G. Mazzanti, S. Bosco, S. Bellucci, A. Cucina, F.L. Foche, G.A. Carru, S. Mastrangelo, A.D. Sotto, R. Masciangelo, A.M. Chiaretti, C. Balasubramanian, G.D. Bellis, F. Micciulla, N. Porta, G. Deriu, A. Tiberia, J. Phys.: Condens. Matter 20 (2008) 1–10. [19] A. Takagi, A. Hirose, T. Nishimura, N. Fukumori, A. Ogata, N. Ohashi, S. Kitajima, S. Kitajima, J. Kanno, J. Toxicol. Sci. 33 (2008) 105–116. [20] T. Thurnherr, C. Brandenberger, K. Fischer, L. Diener, P. Manser, X. MaederAlthaus, J.P. Kaiser, H.F. Krug, B. Rothen-Rutishauser, P.A. Wick, Toxicol. Lett. 200 (2011) 176–186. [21] J.W. Fisher, S. Sarkar, C.F. Buchanan, C.S. Szot, J. Whitney, H.C. Hatcher, S.V. Torti, C.G. Rylander, M.N. Rylander, Cancer Res. 70 (2010) 1–10. [22] Y. Bai, I.S. Park, S.J. Lee, T.S. Bae, F. Watari, M. Uo, M.H. Lee, Carbon 49 (2011) 3663–3671. [23] C.W. Nam, S.J. Kang, Y.K. Kang, M.K. Kwak, Arch. Pharm. Res. 34 (2011) 661– 669. [24] X. Wang, T. Xia, S.A. Ntim, Z. Ji, S. George, H. Meng, H. Zhang, V. Castranova, S. Mitra, A.E. Nel, ACS Nano 4 (2010) 7241–7252. [25] Y. Sakamoto, F.N. Nakae, K. Tayama, A. Maekawa, K. Imai, A. Hirose, T. Nishimura, N. Ohashi, A. Ogata, J. Toxicol. Sci. 34 (2009) 65–76. [26] S. Brahmachari, D. Das, P.K. Das, Chem. Commun. 46 (2010) 8386–8388. [27] M. Horie, M. Stowe, M. Tabei, H. Kato, A. Nakamura, S. Endoh, Y. Morimoto, K. Fujita, Toxicol. Mech. Methods 23 (2013) 315–322. [28] H. Kato, A. Nakamura, M. Horie, S. Endoh, K. Fujita, H. Iwahashi, S. Kinugasa, Carbon 49 (2011) 3989–3997. [29] H. Kato, K. Mizuno, M. Shimada, A. Nakamura, K. Takahashi, K. Hata, S. Kinugasa, Carbon 47 (2009) 3434–3440. [30] A.E. Frise, E. Edri, I. Furó, O. Regev, J. Phys. Chem. Lett. 1 (2010) 1414–1419. [31] A.E. Frise, G. Pagès, M. Shtein, I.P. Bar, O. Regev, I. Furó, J. Phys. Chem. B 116 (2012) 2635–2642. [32] H. Kato, A. Nakamura, M. Horie, RSC Adv. 6 (2014) 2129–2136. [33] D.N. Futaba, K. Hata, T. Namai, T. Yamada, K. Mizuno, Y. Hayamizu, M. Yumura, S. Iijima, J. Phys. Chem. B 110 (2006) 8035–8038. [34] J.E. Tanner, J. Chem. Phys. 52 (1970) 2523–2526. [35] R. Mills, J. Phys. Chem. 80 (1976) 888–890. [36] R.W. O’brien, D.N. Ward, J. Colloid Interface Sci. 121 (1988) 402–413. [37] B.V. Derjaguin, L.D. Landau, Acta Physicochem. URSS 14 (1941) 633–662. [38] E.J.W. Verwey, J.T.H.G. Overbeek, Trans. Faraday Soc. B 42 (1946) 117–123. [39] H. Kato, T. Saito, M. Nabeshima, K. Shimada, S. Kinugasa, J. Magn. Reson. 180 (2006) 266–272. [40] D. Chattopadhyay, L. Galeska, F.A. Papadimitrakopoulos, J. Am. Chem. Soc. 125 (2003) 3370–3375. [41] Z. Chen, S. Nagase, A. Hirsch, R.C. Haddon, T.W. Paul von Ragué Schleyer, Angew. Chem. 116 (2004) 1578–1580. [42] R. Sitko, B. Zawisza, E. Malicka, Trends Anal. Chem. 37 (2012) 22–31.
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Acceleration of suspending single-walled carbon nanotubes in BSA aqueous solution induced by amino acid molecules Haruhisa Kato a,⇑, Ayako Nakamura a, Masanori Horie b a National Metrology Institute of Japan (NMIJ), National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba Central 5, Higashi 1-1-1, Tsukuba, Ibaraki 305-8565, Japan b Health Research Institute (HRI), AIST, Japan
a r t i c l e
i n f o
Article history: Received 26 July 2014 Accepted 9 September 2014 Available online 19 September 2014 Keywords: PFG-NMR SWCNT Dynamic light scattering Amino acid Suspending Dispersibility Zeta potential
a b s t r a c t Single-walled carbon nanotube (SWCNT) suspensions in aqueous media were prepared using bovine serum albumin (BSA) and amino acid molecules. It was found that the amino acid molecules clearly decreased the time required for suspending the SWCNTs in BSA aqueous solutions. Dynamic light scattering measurements revealed that the particle sizes of the SWCNTs suspended in aqueous media with and without amino acid molecules were approximately the same and stable for more than one week. The zeta potential values of the BSA molecules in pure water and amino acid aqueous solutions were different, and these values were also reflected in the surface potential of colloidal SWCNT particles in the corresponding aqueous media, thus inducing different dispersibility of SWCNTs in aqueous media. Pulsed field gradient nuclear magnetic resonance measurements showed that the interactions between the SWCNTs and the amino acid molecules are weak and comprise chemical exchange interactions and not bonding interactions. Amino acid molecules play a fascinating role in the preparation of SWCNT suspensions in BSA aqueous media by increasing electrostatic repulsive interactions between SWCNT colloidal particles and consequently enhancing the dispersion ability of the BSA molecules. Ó 2014 Elsevier Inc. All rights reserved.
1. Introduction Single-walled carbon nanotubes (SWCNTs) have been widely used in chemical and industrial applications, i.e., organic solar cells, transistors, capacitors, drug delivery, and hydrogen storage [1–12]. To utilize SWCNTs for scientific and technological applications, preparation of SWCNT suspensions is a very important step because the bundling structures of SWCNTs strongly affect the electronic structure [13]. However, carbon nanomaterials such as SWCNTs tend to easily self-aggregate through strong van der Waals interactions and hydrophobic interactions in the aqueous phase [14]. Therefore, one of the general techniques for preparing stable SWCNT aqueous suspensions is to suspend SWCNTs using various surfactants such as Triton X, Tween 80, Pluronic F127, sodium dodecyl sulfate, bovine serum albumin (BSA), carboxyl methyl cellulose, and amino acid based amphiphiles under ultrasonication [15–26]. Although chemical functionalization of the SWCNT surface is an effective procedure for suspending SWCNTs in aqueous media individually, surface modification alters the physical and chemical properties of the SWCNTs. Therefore, the ⇑ Corresponding author. Fax: +81 29 861 4618. E-mail address: [email protected] (H. Kato). http://dx.doi.org/10.1016/j.jcis.2014.09.018 0021-9797/Ó 2014 Elsevier Inc. All rights reserved.
suspending procedure for SWCNTs using various surfactants is a very important method. However, appropriate choice of the surfactant is also very important for the preparation of stable and homogeneous aqueous suspensions of SWCNTs. In particular, BSA is a biocompatible dispersant, as shown by in vitro toxicity assessments [27], and works well as an effective dispersant for various nanocarbon materials [28]. However, the time required to suspend SWCNTs in an aqueous phase under ultrasonication is longer than that required for suspending chemically modified SWCNTs in an aqueous phase, resulting in the need for a more effective suspending procedure. The aim of the present study was to develop a procedure for preparing suspensions of SWCNTs within a shorter time using a general biocompatible surfactant, i.e., BSA molecules. The key point of our novel procedure is that we used not only BSA as the dispersant for SWCNTs in the aqueous phase but also amino acid molecules as accelerating agents to enhance the capability of BSA to suspend SWCNTs in aqueous media. The amino acid molecules used in this study are aspartic acid and glutamic acid because they are commercially available and classified as nonessential amino acid molecules. Herein, we discovered a fascinating role of amino acid molecules in the preparation of SWCNT suspensions, and proposed a novel and practical approach that will allow researchers to
H. Kato et al. / Journal of Colloid and Interface Science 437 (2015) 156–162
obtain easily reproducible SWCNT aqueous suspensions using general amino acid molecules. In this study, the dispersibility of the prepared SWCNT suspensions was investigated through dynamic light scattering (DLS) and electrophoretic mobility measurements. The diffusion of amino acid molecules in aqueous media with/without SWCNTs was directly observed by pulsed field gradient nuclear magnetic resonance (PFG-NMR) measurements. Because the PFG-NMR method has the potential to provide a mathematical and physical method for predicting the diffusion phenomena of much smaller targets in solution than DLS, it has been used to study the diffusion phenomena of various surfactant molecules in CNT aqueous suspensions [29–32]; therefore, we employed this method in this study. Using these three measurement methods, we clarified the accelerating effect induced by amino acid molecules on suspending SWCNTs in BSA aqueous solutions, in our novel procedure for the preparation of SWCNT suspensions. 2. Experimental section 2.1. Materials and preparation of aqueous SWCNT suspensions High-purity SWCNT material was synthesized by a waterassisted chemical vapor deposition growth method [33]. The purity and size distribution were carefully evaluated by X-ray Fluorescence spectrometry, transmission electron microscopy observations, and Thermogravimetric Analysis. We used same SWCNT reported by Futaba et al. and the detail properties of this SWCNT are summarized in Ref. [33]. L-aspartic acid (MW = 133.1 g), L-glutamic acid (MW = 147.13 g), and BSA (MW = 66 kDa) were purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). Stable, homogeneous aqueous suspensions of SWCNTs (0.1 mg/mL) were dispersed in D2O aqueous solutions of BSA, L-aspartic acid, and L-glutamic acid using an ultrasonic bath (140 W, 25 kHz). The parameters for the SWCNT/BSA/L-aspartic acid/L-glutamic acid suspensions are summarized in Table 1. 2.2. DLS measurements A DLS7000 (Otsuka Electronics Co., Ltd., Kyoto Japan) with a goniometer system and 45 mW He–Ne laser at a wavelength of 632.8 nm was used. Multiple tau digital correlation was measured at a minimum sampling time of 0.1 ls. The measurements were performed at a scattering angle of 90° with a quartz sample cell. The measurement temperature was regulated at 25.0 °C ± 0.1 °C. The instrument was stored in a clean booth maintained at a constant room temperature of 23.0 °C ± 0.3 °C and humidity of 40% ± 3%. The same sample positions were used for all bottles of SWCNT dispersions for DLS measurements that were stored for longer than one week. Measurements were repeated at least three times, and the mean values were used. Taking the apparent diffusion coefficients obtained by DLS, the hydrodynamic particle sizes of the SWCNT colloidal nanoparticles were determined using the Stokes–Einstein relation, Table 1 Compositions used for SWCNT dispersion. Sample name
SWCNT (mg/mL)
BSA (mg/ mL)
Aspartic acid (mg/mL)
Glutamic acid (mg/mL)
A B1 B2 B3 C1 C2 C3
0.1 – – 0.1 – – 0.1
3.1 – 3.1 3.1 – 3.1 3.1
– 1.2 1.2 1.2 – – –
– – – – 1.2 1.2 1.2
dl ¼
kB T ; 3pgD
157
ð1Þ
where kB is the Boltzmann constant, T is the absolute temperature, g is the viscosity of the solvent, and dl is the calculated hydrodynamic diameter of the colloidal nanoparticles. 2.3. PFG-NMR measurements PFG-NMR measurements were performed on a 14.1 T spectrometer (UNITY INOVA 600A, Varian, California, USA) equipped with an H-F{X} diffusion probe (DSI-V218, Doty Scientific California, USA) capable of producing magnetic field pulse gradients of approximately 2500 G cm1 in the z-direction. The NMR lock was not used for any of these experiments, and the temperature was set at 298.15 K ± 0.1 K. Temperature calibration of the spectrometer was performed using ethylene glycol as a standard. All experiments were carried out using 5 mm o.d. microcell NMR tubes (BMS-005V, Shigemi Co., Ltd., Tokyo, Japan), and the sample height was 5 mm. The diffusion coefficients were calculated using the Stejskal–Tanner diffusion equation,
lnðI=I0 Þ ¼ Dc2 G2 d2 ðD d=3Þ;
ð2Þ
for data with a correlation coefficient of ln(I/I0) vs. higher than 0.99 for the first decay. The PFG stimulated echo (PFGSTE) sequence [(p/2)–s2–(p/2)–s1–(p/2)–s2] acquisition was used [34]. Rectangular pulsed gradients with a duration of 1 ms were incrementally increased from 0 to 595 G cm1 with 64–256 averaged transients. The p/2 pulse width of 12.90 ls, relaxation delay of 10 s, and acquisition time of 2.0 s were defined in advance. The three diffusion times (D) were 50, 75, and 100 ms. Water was purified by a microporous filtration system (Millipore S.A. Japan, Tokyo, Japan) and used as a standard to assure the linearity of the gradient strength from the gradient amplifier. The value used for the water diffusion coefficient was 2.299 109 m2 s1 [35]. 2.4. Electrophoretic mobility measurements Electrophoretic mobility measurements were performed at 25.0 ± 0.1 °C using a zeta potential analyzer (ELSZ, Otsuka Electronics Co., Ltd.). The zeta potentials were estimated in all cases using the simple Smoluchowski equation. Depending on the magnitude of the zeta potential, the spherical Smoluchowski approximation may overestimate the actual zeta potential by up to 20% [36]. 2.5. Raman spectra measurements Raman spectra measurements were performed using a DXR microscope (Thermo Fisher Scientific Inc., Waltham, MA, USA) and recorded with a 532 nm laser line. The SWCNT spectrum presented two peaks at 1580 cm1 and 1340 cm1, which were assigned to G and D bands, respectively. Using these two peaks, the G:D ratios were calculated. Low-energy Raman features that correspond to scattering from the radial breathing mode (RBM) were also used to verify the diameters of SWCNTs. The calculated values were 1.5–1.8 nm. 3. Results and discussion 3.1. Preparation and characterization of SWCNT aqueous colloidal suspensions with and without amino acids The effects of two different dispersants for preparing SWCNT suspensions in aqueous media were investigated as follows. One dispersant was a typical biocompatible dispersant, BSA, and the other dispersant was a mixture of BSA and amino acids. The aqueous
158
H. Kato et al. / Journal of Colloid and Interface Science 437 (2015) 156–162
suspensions of SWCNTs (0.1 mg/mL) were prepared in 3.1 mg/mL aqueous solutions of BSA and BSA/amino acid mixtures (3.1 mg/mL for BSA and 1.2 mg/mL for amino acid molecules) using an ultrasonic bath (140 W, 25 kHz) for 2 h at 15 min intervals. Fig. 1 shows the time required for the preparation of SWCNT suspensions in both the aqueous media. The SWCNT suspension in the aqueous solution of BSA/amino acid (aspartic acid) mixture homogenized and dispersed more rapidly than that in the BSA aqueous solution, in 45 min, as shown in Fig. 1. After 45 min of ultrasonication time, the mean sizes of colloidal SWCNT particles were not changed and took similar/stable sizes. This phenomenon was observed in aqueous solutions of BSA/amino acid mixtures (3.1 mg/mL for BSA and 0.6–1.2 mg/mL for amino acid), however; the required homogenizing time increased at the lower amino acid concentration than 0.6 mg/mL (see Supporting information). In contrast, when SWCNTs were suspended in an aqueous solution with only BSA, more than 100 min was required for the SWCNT aqueous suspension to become homogenous and take stable colloidal size. We also prepared SWCNT aqueous suspensions using glutamic acid instead of aspartic acid and found a similar phenomenon. This result indicates that the amino acid molecules work as accelerators in the preparation of homogenous aqueous suspensions of SWCNTs. DLS measurements were performed for all the SWCNT aqueous suspensions. Although a simple assumption was used (e.g. size distribution, sphere structure. . .) in the DLS analysis for SWCNT colloidal particles, the monitoring of the apparent diameter of colloidal particles by DLS over a period is useful to examine the stability of prepared SWCNT aqueous suspensions and to compare the sizes among suspensions. The hydrodynamic diameters, expressed as light scattering intensity-averaged diameters, are shown in Fig. 2. We used a reverse Laplace transformation for light scattering data analysis because the raw data agreed with the observed photon correlation function, indicating that the SWCNT samples followed an approximate monomodal size distribution.
BSA + amino acid
As shown in Fig. 2, the values obtained from the DLS measurements for SWCNT/BSA and SWCNT/BSA/amino acid suspensions were unchanged for more than one week after preparation. These results indicated that BSA and BSA/amino acid worked well as dispersants for SWCNTs in aqueous media, and the observed hydrodynamic sizes of the SWCNTs were similar for all suspensions, regardless of whether BSA or BSA/amino acid was used. Amino acid molecules worked as accelerating agents that enhanced the capability of BSA to suspend SWCNTs in aqueous media. The zeta potentials of the SWCNT aqueous suspensions were measured to estimate changes in the surface potentials depending on the different dispersant combinations. The uncertainties in the table were calculated from the repeatability of the observed zeta potentials, which were obtained from at least three separate measurements. The zeta potentials of the SWCNTs in the BSA aqueous solution and in aqueous solutions of BSA/amino acid (aspartic acid and glutamic acid) mixtures were obviously different. As shown in Table 2, the zeta potential of BSA molecules was approximately 18 mV and that of SWCNT colloidal particles in BSA aqueous solution (A) was 22 mV, indicating that the BSA molecules were adsorbed on the SWCNTs. However, the estimated value of the zeta potential of SWCNTs in BSA/amino acid aqueous solutions (B3 and C3) was 30 mV for both kinds of amino acid molecules. The zeta potentials of BSA molecules in both amino acid aqueous solutions (B2 and C2) were measured, and the values are shown in Table 2. Interestingly, the zeta potentials of B2 and C2 were similar to those of B3 and C3. The zeta potentials of BSA molecules in amino acid aqueous solutions increased, and the absolute values became larger upon the addition of amino acid molecules, indicating that the amino acid molecules interacted on the surfaces of BSA molecules. This result showed that the surface-modified BSA molecules work as excellent dispersants of SWCNTs in aqueous phases. The zeta potentials of SWCNTs in aqueous solutions of BSA/amino acid (B3 and C3) reflected those of the BSA molecules whose surface potential was modified by amino acids (B2 and C2). According to
BSA
0 min
45 min
90 min
15 min
60 min
105 min
30 min
75 min
120 min
Fig. 1. Photographs of SWCNT suspensions: left figures show SWCNT/BSA/amino acid (aspartic acid) aqueous suspensions, and right figures show SWCNT/BSA aqueous suspensions. Faster dispersion and homogenization, in 45 min, were observed in the SWCNT/BSA/amino acid aqueous suspension, whereas more than 100 min was required to obtain the homogenous SWCNT/BSA aqueous suspension.
159
H. Kato et al. / Journal of Colloid and Interface Science 437 (2015) 156–162 Table 3 G:D ratios of the SWCNTs in various suspensions.
A B3 C3
Sample name
GD ratio
Standard uncertainty
A B3 C3
4.2 4.6 4.0
0.3 0.6 0.2
In Table 3, the G:D ratios for all SWCNT suspensions were calculated from Raman spectroscopic measurements. We could not find any obvious difference between the values for the three suspensions, indicating that the amino acid molecules accelerated the suspending process and did not chemically react to decompose the SWCNTs in the aqueous phases. Fig. 2. Plots of hydrodynamic diameter of SWCNTs vs. time (d). The diameter did not change during the measurement, indicating no agglomeration for SWCNT/BSA and SWCNT/BSA/amino acid suspensions.
Table 2 Zeta potentials of the SWCNT suspensions. Sample name
Zeta potential (mV)
Standard uncertainty (mV)
A B2 B3 C2 C3 BSA
21.5 34.0 29.6 26.5 30.4 18.2
0.6 2.0 1.0 1.4 1.4 1.4
the DLVO (Derjaguin, Landau, Verwey, and Overbeek) theory [37,38], the dispersibility of a material suspension is based on two different interactions, electrostatic repulsive interactions and van der Waals attractive interactions. Since the sizes of colloidal SWCNT particles both in BSA aqueous solution and BSA/amino acid aqueous solution were same as shown in Fig. 2, it was assumed that the van der Waals attractive interactions of A, B3, and C3 were approximately similar. Herein, we therefore focused on the electrostatic repulsive interaction between colloidal SWCNT particles. The electrostatic repulsive energy is not equal to the zeta potential because the latter is the potential for the slip face where a liquid starts to flow inside the electrical double layer formed on the hydrodynamic surface of a material. However, these two values are strongly related to each other. Thus, the absolute values of the zeta potential were larger in B3 than in A, which indicated that the electric repulsive interaction between SWCNT colloidal particles was caused by the amino acid molecules that interacted with the surfaces of the SWCNTs. According to DLVO theory, the repulsive interaction between nanoparticles from the electrical doublelayer can be described by
VðRÞ ¼ pe0 edW2 ejr :
3.2. Characterization of amino acid molecules in aqueous solution/ SWCNT suspensions by PFG-NMR Fig. 3 shows the 1H spectra of the various aqueous solutions/ suspensions (A, B1, B2, and B3) measured with the PFGSTE pulse sequence at a gradient strength of 0 G cm1 and 50 G cm1. The peaks of the BSA molecules broadened and those of the amino acids were sharp. The sharp methine peak (NH2CH) at 3.9–3.7 ppm was close to the water peak, and the sharp peaks from 3.1 to 2.6 ppm were assigned as the methylene protons (CH2). The spectral peak at 3.9–3.7 ppm was sufficiently strong, as shown in the figure, allowing us to carry out PFG-NMR diffusion measurements using this peak. To determine the appropriate parameters for the pulse sequence, we measured the longitudinal (T1) and transverse (T2)
(a)
(b)
ð3Þ
In the above expressions, e0 is the permittivity of vacuum, e is the dielectric constant of the electrolyte solution, e is the elemental charge, kB is the Boltzmann constant, T is the absolute temperature, q/i is the ion number density in the bulk electrolyte, z is the valence of ion i, and j is the inverse of the Debye length. Assuming that the zeta potentials are equal to the surface potentials of SWCNT colloidal particles in Eq. (3), the electrical repulsive energy of B3 and C3 (the zeta potentials were approximately 30 mV) was therefore approximately 2.3 times larger than that of the SWCNTs in the just BSA aqueous solution (the zeta potentials were approximately 20 mV) due to V(R) / W2 in this case. It was estimated that the increased surface charge of the SWCNTs in the suspension in samples B3 and C3 was a significant factor in decreasing the suspending time of the SWCNTs in the aqueous phase.
Fig. 3. Examples of 1H NMR spectra of surfactant solutions collected using the PFGSTE pulse sequence at gradient strengths of 0 G cm1 and 50 G cm1. The diffusion time was 50 ms. (a) A (black), B1 (red), B2 (green), and B3 (blue) at a gradient strength of 0 G cm1; (b) A (black), B1 (red), B2 (green), and B3 (blue) at a gradient strength of 50 G cm1. The peak at 4.7 ppm was assigned to HDO. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
160
H. Kato et al. / Journal of Colloid and Interface Science 437 (2015) 156–162
Table 4 Longitudinal (T1) and transverse (T2) relaxation times of amino acid molecules at 3.8 ppm. Sample name
T1 (s)
T2 (s)
B1 B2 B3 C1 C2 C3
4.6 4.2 3.3 4.5 4.0 3.1
1.3 1.2 1.0 1.2 1.1 0.8
relaxation times at 3.9–3.7 ppm for the typical peak for both amino acids, as shown in Fig. 3. The observed results are summarized in Table 4. The values of T1 and T2 were similar for both amino acid molecules. For these PFG-NMR measurements, the recycle decays were set at 10 s, and the observed diffusion times were set at 50, 75, and 100 ms. The T1 and T2 values slightly changed between B1 vs. B2 and C1 vs. C2, indicating that the BSA molecules and amino acid molecules interacted weakly with each other, since the relaxation value is strongly influenced by slow molecular motion. A larger reduction of the T1 and T2 values was found in the SWCNT/BSA/amino acid suspensions (B3 and C3) than in the BSA/amino acid solutions (B2 and C2), indicating that the amino acid molecules interacted more strongly with the SWCNT/BSA colloidal particles than the BSA molecules. However, the observed differences in the relaxation times among B1/B2/B3 and C1/C2/C3 were small (10–20% difference), and the amino acid molecules individually diffused into the solutions/suspensions. The plots obtained from the PFGSTE echo signal attenuation for the two amino acid molecules are shown in Fig. 4. The attenuation plots were approximately linear and did not depend on the diffu-
sion times, indicating that the two amino acid molecules completely dissolved in water. The diffusion coefficients of the amino acid molecules calculated using linear regression (solid lines, Fig. 4) were 7.34 ± 0.09 1010 and 6.64 ± 0.05 1010 m2 s1 for samples B1 and C1, respectively. The uncertainties in the diffusion coefficients were calculated according to Ref. [39]. The small difference in the diffusion coefficients was induced by the difference in the length of the methylene unit between the two amino acid molecules. The plots obtained from the PFGSTE echo signal attenuation for the two amino acid molecules in BSA aqueous solutions are shown in Fig. 5. The attenuation plots were approximately linear and did not depend on the diffusion times, indicating that the interaction between amino acid and BSA molecules is weak. The diffusion coefficients of the amino acid molecules calculated by linear regression (solid lines, Fig. 5) were 6.75 ± 0.13 1010 and 6.01 ± 0.07 1010 m2 s1 for samples B2 and C2, respectively. The calculated diffusion coefficients of amino acid molecules for B2 and C2 were slightly lower than the corresponding values for B1 and C1, respectively. In addition, as shown in Table 4, the relaxation times for two amino acid molecules were slightly reduced. These results show that the BSA and amino acid molecules interacted weakly and were not strongly bonded to each other, since there was no attenuation peak that agreed with the diffusion coefficient of 7 1011 m2 s1, which is the value for BSA molecules, in the PFG-NMR measurements. Fig. 6 shows the plots of the PFGSTE echo signal attenuation for the two amino acid molecules in BSA/SWCNT aqueous suspensions. Interestingly, the attenuation plots were still approximately linear but did not show the same dependency on the diffusion times. The values of the diffusion coefficients of the amino acid molecules calculated by linear regression (solid lines, Fig. 6) were
(a)
γ
δ Δ δ/3
(a)
γ
δ Δ δ/3
(b)
(b)
γ
δ Δ δ/3
Fig. 4. PFG-NMR spin-echo signal attenuation plots for amino acid aqueous solutions (a) B1 and (b) C1 at d = 1 ms for diffusion times of D = 50 ms (d), D = 75 ms (j), D = 100 ms (N). The solid line represents the linear regression. The diffusion coefficients calculated from the slope are 7.34 ± 0.09 1010 and 6.64 ± 0.05 1010 m2 s1 for B1 and C1, respectively.
γ
δ Δ δ/3
Fig. 5. PFG-NMR spin-echo signal attenuation plots for amino acid/BSA aqueous solutions (a) B2 and (b) C2 at d = 1 ms for diffusion times of D = 50 ms (d), D = 75 ms (j), D = 100 ms (N). The solid line represents the linear regression. The diffusion coefficients calculated from the slope are 6.75 ± 0.13 1010 and 6.01 ± 0.07 1010 m2 s1 for B2 and C2, respectively.
161
H. Kato et al. / Journal of Colloid and Interface Science 437 (2015) 156–162
(a)
Table 5 Diffusion coefficients of aspartic acid and glutamic acid molecules obtained from PFGNMR. Sample name B1 B2 B3 C1 C2 C3
γ
δ Δ δ/3
(b)
γ
δ Δ δ/3
Fig. 6. PFG-NMR spin-echo signal attenuation plots for amino acid/BSA/SWCNT aqueous suspensions (a) B3 and (b) C3 at d = 1 ms for diffusion times of D = 50 ms (d), D = 75 ms (j), D = 100 ms (N). The solid line represents the linear regression. The diffusion coefficients calculated from the slope are 6.51 ± 0.11 1010 and 5.46 ± 0.11 1010 m2 s1 for B3 and C3, respectively.
6.51 ± 0.11 1010 and 5.46 ± 0.11 1010 m2 s1 for samples B3 and C3, respectively. Although very low diffusion PFGSTE signal attenuation was expected because of the strongly adsorbed amino acid molecules on/in the SWCNT colloidal particles, only slightly slower diffusion of the amino acid molecules was observed as compared to those in water or BSA aqueous solutions. As shown in Table 4, a reduction in the relaxation times for the two amino acid molecules is observed in SWCNT suspensions (B3 and C3). However, the changes are not large, thereby supporting the slight changes in the diffusion coefficients of the amino acid molecules (Fig. 6). These results indicated that the amino acid molecules and colloidal SWCNT particles interacted weakly, for example, through a chemical exchange on the surface of the colloidal particles, and that they do not aggregate/agglomerate in aqueous suspensions. Table 5 summarizes the observed diffusion coefficients for the two amino acid molecules under various conditions. A reduction in the diffusion coefficients of the amino acid molecules in the presence of BSA molecules or SWCNT colloidal particles could be clearly seen; however, this reduction was only 10–20%. It is clear that amino acid molecules reduce the suspending time of SWCNTs in BSA aqueous solutions, as shown in Fig. 1. The interaction between the SWCNT colloidal particles and the amino acid molecules is very weak and probably similar to a chemical exchange. Nevertheless, it was observed that the suspending speed of SWCNTs in BSA aqueous solutions was obviously higher with the amino acid molecules than without the amino acid molecules, indicating that surface potential modification of the BSA molecules by the amino acid molecules is a significant force for effectively suspending SWCNTs. This modification results from the increased electrostatic repulsive interactions between SWCNT
Diffusion coefficient (m2 s1) 10
7.34 10 6.75 1010 6.51 1010 6.64 1010 6.01 1010 5.46 1010
Standard uncertainty (m2 s1) 8.91 1012 1.33 1011 1.08 1011 4.76 1012 6.47 1012 1.09 1011
colloidal particles, as observed from zeta potential measurements. Over this decade, surface chemical modifications of SWCNTs have been performed to effectively suspend SWCNTs in aqueous phases [40–42]. However, in this study, it has been shown that surface potential changes of the surfactants are also an important aspect in preparing stable and homogeneous aqueous suspensions of SWCNTs. We could not observe obvious difference in the acceleration effects between aspartic acid and glutamic acid molecules for suspending SWCNTs in BSA aqueous solution, because of the different lengths of the methylene units and the numbers of –COOH group; however, the observed diffusion coefficients were slightly different because of the difference in the molecular weights, as shown in Table 5. Finally, we believe that enhancing the dispersing capability of the surfactant BSA for effectively suspending carbon nanomaterials in aqueous media by means of the increased electrostatic repulsive interactions between the nanomaterials has significant applications in the industrial field. In addition, for in vivo toxicity assessments of SWCNTs, SWCNT bundle aqueous suspensions with various types of surfactants are typically used. However, according to this study, amino acid molecules in the body can be expected to work as accelerators for re-suspending the SWCNT bundles inside the body. This finding is significant in the consideration of the clearance of not only SWCNTs but also other nanomaterials for in vivo toxicity assessments.
4. Conclusions The acceleration effects of amino acid molecules on the preparation of SWCNT suspensions in BSA aqueous solutions were investigated. The time required for suspending SWCNTs while using an ultrasonic bath with amino acid/BSA aqueous solution was clearly shorter than that for the BSA aqueous solution. The zeta potentials of BSA molecules in pure water and amino acid aqueous solutions were different, approximately 20 mV and 30 mV, respectively. These zeta potentials were similar to the surface potential of SWCNT colloidal particles in the corresponding aqueous media, indicating that modification of the surface potential of BSA by amino acid molecules induces differences in the dispersibility of SWCNTs in aqueous media based on electrostatic repulsive interactions. It was also found that the interactions between the SWCNTs and amino acid molecules are very weak and may be a chemical exchange interaction, and not a binding interaction on/in the SWCNTs. Our studies provide significant findings that are expected to be useful for nanomaterial-related research in industrial and biological fields.
Acknowledgments This work was funded by the New Energy and Industrial Technology Development Organization of Japan (NEDO) Grant ‘‘Evaluating risks associated with manufactured nanomaterials (P06041).’’
162
H. Kato et al. / Journal of Colloid and Interface Science 437 (2015) 156–162
Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcis.2014.09.018. References [1] T. Chen, S. Wang, Z. Yang, Q. Feng, X. Sun, L. Li, Z. Wang, Angew. Chem., Int. Ed. 50 (2011) 1815–1819. [2] A. Bianco, K. Kostarelos, C.D. Partidos, M. Prato, Chem. Commun. 57 (2005) 571–577. [3] H. Yoon, J. Ahn, P. Barone, K. Yum, R. Sharma, A. Boghosian, J.H. Han, M.S. Strano, Angew. Chem., Int. Ed. 50 (2011) 1828–1831. [4] J.A. Fagan, M.L. Becker, J. Chun, E.K. Hobbie, Adv. Mater. 20 (2008) 1609–1613. [5] M.A. Bissett, I. Köper, J.S. Quinton, J.G. Shapter, Phys. Chem. Chem. Phys. 13 (2011) 6059–6064. [6] C. Vijiayakumar, B. Balan, M. Kim, M. Takeuchi, J. Phys. Chem. 115 (2011) 4533–4539. [7] Y. Zhao, L. Hu, J.F. Stoddart, G. Grüner, Adv. Mater. 20 (2008) 1910–1915. [8] X. Chen, A. Kis, A. Zettl, C.R. Bertozzi, Proc. Natl. Acad. Sci. 104 (2007) 8218– 8222. [9] A. Rajan, M.S. Strano, D.A. Heller, T. Hertel, K. Schulten, J. Phys. Chem. B 112 (2008) 6211–6213. [10] A. Mateo-Alonso, C. Ehli, K.H. Chen, D.M. Guldi, M. Prato, J. Phys. Chem. A 111 (2007) 12669–12673. [11] F. Ding, P. Larsson, J.A. Larsson, R. Ahuja, H. Duan, A. Rosen, K. Bolton, Nano Lett. 8 (2008) 463–468. [12] R. Marquis, C. Greco, I. Sadokierska, S. Lebedkin, M.M. Kappes, T. Michel, L. Alvarez, J.L. Sauvajol, S. Meunier, C. Mioskowski, Nano Lett. 8 (2008) 1830– 1835. [13] H. Ozawa, T. Fujigaya, Y. Niidome, N. Hotta, M. Fujiki, N. Nakashima, J. Am. Chem. Soc. 133 (2011) 2651–2657. [14] M.J. O’Connell, S.M. Bachilo, C.B. Huffman, V.C. Moore, M.S. Strano, E.H. Haroz, K.L. Rialon, P.J. Boul, W.H. Noon, C. Kittrell, J. Ma, R.H. Hauge, R.B. Weisman, Science 297 (2002) 563–566. [15] A.L. Alpatova, W. Shan, P. Babica, B.L. Upham, A.R. Rogensues, S.J. Masten, E. Drown, A.K. Mohanty, E.C. Alocilja, V.V. Tarabara, Water Res. 44 (2010) 505– 520. [16] H.N. Yehia, R.K. Draper, C. Mikoryak, R.K. Walker, P. Bajaj, I.H. Musselman, M.C. Daigrepont, G.R. Dieckmann, P. Pantano, J. Nanobiotechnol. 5 (2007) 1–17. [17] P. Wick, P. Manser, L.K. Limbach, U. Dettlaff-Weglikowska, F. Krumeich, S. Roth, W.J. Stark, A. Bruinink, Toxicol. Lett. 168 (2007) 121–131.
[18] M. Chiaretti, G. Mazzanti, S. Bosco, S. Bellucci, A. Cucina, F.L. Foche, G.A. Carru, S. Mastrangelo, A.D. Sotto, R. Masciangelo, A.M. Chiaretti, C. Balasubramanian, G.D. Bellis, F. Micciulla, N. Porta, G. Deriu, A. Tiberia, J. Phys.: Condens. Matter 20 (2008) 1–10. [19] A. Takagi, A. Hirose, T. Nishimura, N. Fukumori, A. Ogata, N. Ohashi, S. Kitajima, S. Kitajima, J. Kanno, J. Toxicol. Sci. 33 (2008) 105–116. [20] T. Thurnherr, C. Brandenberger, K. Fischer, L. Diener, P. Manser, X. MaederAlthaus, J.P. Kaiser, H.F. Krug, B. Rothen-Rutishauser, P.A. Wick, Toxicol. Lett. 200 (2011) 176–186. [21] J.W. Fisher, S. Sarkar, C.F. Buchanan, C.S. Szot, J. Whitney, H.C. Hatcher, S.V. Torti, C.G. Rylander, M.N. Rylander, Cancer Res. 70 (2010) 1–10. [22] Y. Bai, I.S. Park, S.J. Lee, T.S. Bae, F. Watari, M. Uo, M.H. Lee, Carbon 49 (2011) 3663–3671. [23] C.W. Nam, S.J. Kang, Y.K. Kang, M.K. Kwak, Arch. Pharm. Res. 34 (2011) 661– 669. [24] X. Wang, T. Xia, S.A. Ntim, Z. Ji, S. George, H. Meng, H. Zhang, V. Castranova, S. Mitra, A.E. Nel, ACS Nano 4 (2010) 7241–7252. [25] Y. Sakamoto, F.N. Nakae, K. Tayama, A. Maekawa, K. Imai, A. Hirose, T. Nishimura, N. Ohashi, A. Ogata, J. Toxicol. Sci. 34 (2009) 65–76. [26] S. Brahmachari, D. Das, P.K. Das, Chem. Commun. 46 (2010) 8386–8388. [27] M. Horie, M. Stowe, M. Tabei, H. Kato, A. Nakamura, S. Endoh, Y. Morimoto, K. Fujita, Toxicol. Mech. Methods 23 (2013) 315–322. [28] H. Kato, A. Nakamura, M. Horie, S. Endoh, K. Fujita, H. Iwahashi, S. Kinugasa, Carbon 49 (2011) 3989–3997. [29] H. Kato, K. Mizuno, M. Shimada, A. Nakamura, K. Takahashi, K. Hata, S. Kinugasa, Carbon 47 (2009) 3434–3440. [30] A.E. Frise, E. Edri, I. Furó, O. Regev, J. Phys. Chem. Lett. 1 (2010) 1414–1419. [31] A.E. Frise, G. Pagès, M. Shtein, I.P. Bar, O. Regev, I. Furó, J. Phys. Chem. B 116 (2012) 2635–2642. [32] H. Kato, A. Nakamura, M. Horie, RSC Adv. 6 (2014) 2129–2136. [33] D.N. Futaba, K. Hata, T. Namai, T. Yamada, K. Mizuno, Y. Hayamizu, M. Yumura, S. Iijima, J. Phys. Chem. B 110 (2006) 8035–8038. [34] J.E. Tanner, J. Chem. Phys. 52 (1970) 2523–2526. [35] R. Mills, J. Phys. Chem. 80 (1976) 888–890. [36] R.W. O’brien, D.N. Ward, J. Colloid Interface Sci. 121 (1988) 402–413. [37] B.V. Derjaguin, L.D. Landau, Acta Physicochem. URSS 14 (1941) 633–662. [38] E.J.W. Verwey, J.T.H.G. Overbeek, Trans. Faraday Soc. B 42 (1946) 117–123. [39] H. Kato, T. Saito, M. Nabeshima, K. Shimada, S. Kinugasa, J. Magn. Reson. 180 (2006) 266–272. [40] D. Chattopadhyay, L. Galeska, F.A. Papadimitrakopoulos, J. Am. Chem. Soc. 125 (2003) 3370–3375. [41] Z. Chen, S. Nagase, A. Hirsch, R.C. Haddon, T.W. Paul von Ragué Schleyer, Angew. Chem. 116 (2004) 1578–1580. [42] R. Sitko, B. Zawisza, E. Malicka, Trends Anal. Chem. 37 (2012) 22–31.
