Anal. Chem. 2010, 82, 5815–5818
Microthermogravimetry of a Single Microcapsule Using Silicon Microresonators Dongkyu Lee,† Yongbeom Park,† Soo Hyoun Cho,‡ Myungsun Yoo,† Namchul Jung,† Minhyuk Yun,† Wooree Ko,† and Sangmin Jeon*,† Department of Chemical Engineering, Pohang University of Science and Technology, Pohang, Korea, and Technical Research Laboratory, POSCO, 699, Gumho-Dong, Gwangyang, Jeonnam, 549-090, South Korea A chlorobenzene-containing polyurethane microcapsule was placed on the free end of a silicon cantilever, and the temperature dependence of the resonance frequency was measured. As the cantilever was heated, the resonance frequency showed steplike increases at 109 and 270 °C that were due to the rupture of the capsule and the thermal degradation of the polyurethane shell, respectively. The frequency changes due to the rupture of a single capsule measured by the cantilever were much sharper than the transitions measured by conventional thermogravimetric analysis (TGA), which measures the average mass change of a collection of capsules characterized by a large size distribution. When two capsules were placed on the cantilever, their individual rupture temperatures could be clearly identified. In addition, the permeability of the polyurethane shell, with respect to chlorobenzene, was measured, and the rupture temperature was observed to decrease with increasing permeability. Microencapsulation techniques have been developed to preserve enclosed materials (for example, in isolating vitamins from oxygen,1 preventing the evaporation of volatile liquids,2,3 and controlling the release of drugs or nutrients4-6). In addition to these traditional applications, microencapsulation has attracted much attention as a basic component of self-healing smart materials.7-9 The simplest self-healing materials are designed using polymer composites containing microencapsulated solvents in the active phase. When the composite is damaged by external forces, the solvent is released from the broken microcapsules and * Author to whom correspondence should be addressed. E-mail: jeons@ postech.ac.kr. † Department of Chemical Engineering, Pohang University of Science and Technology. ‡ Technical Research Laboratory, POSCO. (1) Desai, K. G. H.; Park, H. J. J. Microencap. 2005, 22, 179–192. (2) Yow, H. N.; Routh, A. F. Soft Matter 2006, 2, 940–949. (3) Haefliger, O. P.; Jeckelmann, N.; Ouali, L.; Leo´n, G. Anal. Chem. 2010, 82, 729–737. (4) MA, Y.; Dong, W. M.; Hempenius, A.; Mo ¨hwald, H.; Vancso, G. J. Nat. Mater. 2006, 5, 724–729. (5) Wang, K.; He, Q.; Yan, X.; Cui, Y.; Qi, W.; Duan, L.; Li, J. J. Mater. Chem. 2007, 17, 4018–4021. (6) Wang, X.; Xie, X.; Cai, C.; Rytting, E.; Steele, T.; Kissel, T. Macromolecules 2008, 41, 2791–2799. (7) Caruso, M. M.; Delafuente, D. A.; Ho, V.; Sottos, N. R.; Moore, J. S.; White, S. R. Macromolecules 2008, 40, 8830–8832. (8) White, S. R.; Sottos, N. R.; Geubelle, P. H.; Moore, J. S.; Kessler, M. R.; Sriram, S. R.; Brown, E. N.; Viswanathan, S. Nature 2001, 409, 794–797. (9) Caruso, M. M.; Blaiszik, B. J.; White, S. R.; Sottos, N. R.; Moore, J. S. Adv. Funct. Mater. 2008, 18, 1898–1904. 10.1021/ac100913k 2010 American Chemical Society Published on Web 06/03/2010
dissolves the surrounding polymer matrix to repair the damaged region. Because the preparation of such polymer composites generally requires thermal curing steps, it is important to investigate the thermal properties of the impregnated microcapsules. The thermal characteristics of microcapsules are usually measured using well-established methods, such as differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA).10-14 TGA, in particular, is routinely used to analyze composition, reaction kinetics, thermal stability, and the thermomechanical properties of microcapsules.13,14 However, determination of an accurate core:shell mass ratio becomes problematic in the case of weak microcapsules that do not survive the agitation process used to prepare them, which can result in a mixture of ruptured and intact microcapsules in the TGA sample. The synthesis of smaller capsules requires faster agitation or ultrasonic cavitation, which breaks even more capsules and results in large TGA measurement errors. This problem can be avoided by measuring the thermal properties of single microcapsules. However, conventional TGA measurements require a collection of microcapsules, because the method can only detect mass changes on the microgram scale. In contrast, microcantilevers can typically detect picogram mass changes. In addition, fast thermal equilibration and multiple sample measurements are possible because of the miniaturized array structure. Despite the obvious advantages of cantilever sensors as a tool for investigating the thermal properties of single microcapsules, most such studies have focused on the detection of gases and biomolecules,15-17 and only a few studies have used cantilever sensors for thermogravimetric measurements.18-20 In this study, we synthesized polyurethane microcapsules containing chlorobenzene and measured their thermomechanical properties using silicon microcantilevers. To our knowledge, this is the first such study to be reported. Variations in the resonance (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20)
Torini, L.; Argillier, J. F.; Zydowicz, N. Macromolecules 2005, 38, 3225–3236. Park, S.; Shin, Y.; Lee, J.-R. J. Colloid Interface Sci. 2001, 241, 502–508. Vyazovkin, S. Anal. Chem. 2008, 80, 4301–4316. Cho, S. H.; White, S. R.; Braun, P. V. Adv. Mater. 2009, 21, 645–649. Croll, L. M.; Sto1o ¨ver, H. D. H.; Hitchcock, A. P. Macromolecules 2005, 38, 2903–2910. Fritz, J.; Baller, M. K.; Lang, H. P.; Rothuizen, H.; Vettiger, P.; Meyer, E.; Guntherodt, H. J.; Gerber, C.; Gimzewski, J. K. Science 2000, 288, 316–318. Thundat, T.; Chen, G. Y.; Warmack, R. J.; Allison, D. P.; Wachter, E. A. Anal. Chem. 1995, 67, 519–521. Krause, A. R.; Neste, C.; Senesac, L.; Thundat, T.; Finot, E. J. Appl. Phys. 2008, 103, 094906. Berger, R.; Lang, H. P.; Gerver, C.; Gimzewski, J. K.; Fabian, J. H.; Scandella, L.; Meyer, E.; Gntherodt, H.-J. Chem. Phys. Lett. 1998, 294, 363–369. Lee, J.; King, W. P. Rev. Sci. Instrum. 2008, 79, 054901. Ono, T.; Esashi, M. Meas. Sci. Technol. 2008, 15, 1977–1981.
Analytical Chemistry, Vol. 82, No. 13, July 1, 2010
5815
frequency of the microcapsule-mounted cantilever result from mass changes of the capsule; therefore, the permeability of chlorobenzene through the polyurethane shell, as well as the rupture temperature of the capsule, can be obtained from resonance frequency measurements. EXPERIMENTAL SECTION Materials. Cyclohexanone, tolylene-2,4-diisocyanate, and 1,4butanediol were used to synthesize urethane prepolymers. Gum arabic from the acacia tree and ethylene glycol were used as the surfactant and cross-linking agent, respectively. Chlorobenzene is commonly adopted for solvent-promoted self-healing capsules and was therefore used as the core material. All chemicals were purchased from Aldrich (Saint Louis, MO) and used as-received. Rectangular silicon cantilevers were obtained from Micromotive (Mainz, Germany). As stated by the manufacturer, each cantilever was 450 µm long, 90 µm wide, and 900 nm thick, with a spring constant of 0.06 N/m. Synthesis of Microcapsules. The urethane prepolymer was synthesized by mixing 21.85 g of tolylene-2,4-diisocyanate and 141.65 g of cyclohexanone in a round-bottomed flask. A 4.12 g (46 mmol) sample of 1,4-butanediol was slowly added to the mixture under agitation at 300 rpm. The solution was reacted for 24 h at 80 °C. After evaporation of the cyclohexanone under reduced pressure at 100 °C, a yellowish and viscous prepolymer was obtained. Microcapsules were synthesized via in situ interfacial polymerization, as described elsewhere.13 In brief, 2.91 g of the synthesized urethane prepolymer were mixed with 12.5 g of chlorobenzene. After dissolution of the prepolymer in the solvent, the mixture was added to the 15 wt % gum arabic aqueous solution (28.0 g), and 1.5 g (24 mmol) of ethylene glycol were slowly added at 50 °C. After 2 h of heating at 70 °C with stirring at 2000 rpm, the solvent-encapsulated microcapsules were synthesized. Figures 1a and 1b show an optical microscopy image of the synthesized microcapsules and a cross-sectional scanning electron microscopy (SEM) image of a broken capsule, respectively. Figure 1c shows the size distribution of the synthesized capsules, which displayed a mean size of ∼55 µm. Measurements of the Temperature-Dependent Mass Changes of Microcapsules. A silicon cantilever was mounted onto a thin aluminum holder, and the temperature was controlled using a resistance heater with a programmable temperature controller (Hanyoung, Inchon, Korea). A sharpened glass pipet was used to mount the single microcapsule on the cantilever apex where the resonance frequency of the cantilever is not affected by the modulus change of the capsule.21,22 The temperature was linearly increased from 50 °C to 350 °C at a rate of 2 °C/min. The motion of the microcantilever was tracked using a laser beam reflected from the cantilever surface onto a duo-lateral positionsensitive detector (SiTek Electro Optics, Partille, Sweden). The resonance frequency of the thermally vibrating cantilever was measured using a fast Fourier transform (FFT) algorithm. A control experiment showed that the resonance frequency of the silicon cantilever decreased by ∼20 Hz with the temperature ranging from 50 °C to 350 °C, because of the decrease in the (21) Lee, D.; Kim, S.; Jung, N.; Thundat, T.; Jeon, S. J. Appl. Phys. 2009, 106, 024310. (22) Dohn, S.; Sandberg, R.; Svendsen, W.; Boisen, A. Appl. Phys. Lett. 2005, 86, 233501.
5816
Analytical Chemistry, Vol. 82, No. 13, July 1, 2010
Figure 1. (a) Optical microscopy image of chlorobenzene-containing microcapsules when synthesized at a stirring rate of 2000 rpm. (b) Cross-sectional scanning electron microscopy (SEM) image of a broken capsule. (c) Size distribution of the synthesized capsules.
modulus of silicon, which is much smaller than that which is due to the evaporation of chlorobenzene from the microcapsule (>1 kHz). (See Figure S1 in the Supporting Information.) The images of a capsule-mounted cantilever were captured in situ using an optical microscope during the experiment. The resonance frequency of a single capsule-mounted rectangular cantilever is given by19
f)
1 2π
k 0.24mc + mm
(1)
where k is the spring constant, and mc and mm are the masses of the cantilever and microcapsule, respectively. Note that the resonance frequency is affected by both the mass and spring constant of the cantilever. The temperature dependence of the resonance frequency can be obtained by differentiating eq 1 with respect to temperature:18,19 dmm dk df ) f dT 2k dT 2(0.24mc + mm) dT
(2)
Thus, the variation in the mass of the mounted capsule, as a function of temperature, can be calculated using
(
mm(T) ) 0.24mc
f0(T)2 f(T)2
)
-1
(3)
where f0 is the resonance frequency of the cantilever without the capsule at a given temperature T.18,19
Figure 3. Temperature-dependent variations in the weight percentage of the microcapsules measured using the commercial TGA (black trace) and the microcantilevers (other color traces).
Figure 2. (a) Temperature-dependent variations in the resonance frequency (black trace) and the first derivative of the frequency (blue trace) of a single capsule-loaded cantilever. (b) Variations in the mass of a single capsule as a function of temperature. The insets show in situ optical microscopy images of the microcapsule-mounted cantilever at 50, 109, and 350 °C.
RESULTS AND DISCUSSION Figure 2a shows the variations in the resonance frequency of a single capsule-mounted cantilever and the first derivative of these frequency variations during heating. As the temperature increased from 50 °C, the resonance frequency of the cantilever increased gradually, because of the evaporation of the solvent through the polyurethane shell. An abrupt increase in the resonance frequency was observed at 109 °C, which corresponded to the rupture temperature of the capsule. The rupture temperature of the microcapsule was determined from the first derivative of the frequency, with respect to temperature. Further heating to 250 °C induced negligible changes in the resonance frequency. Additional steep changes in the frequency were observed at 270 °C, because of thermal degradation of the polyurethane shell. Figure 2b shows the temperature-dependent mass of a single microcapsule, calculated using eq 3. The initial mass of the capsule was 175 ng, and this mass decreased to 30 ng after complete evaporation of chlorobenzene, indicating that the capsule consisted of 82 wt % chlorobenzene and 18 wt % polyurethane shell. Considering that the density of polyurethane is ∼1.1 g/cm3 and the diameter of the capsule is ∼70 µm, the thickness of the polyurethane shell was calculated to be ∼1.8 µm, which is a
typical polyurethane shell thickness, as determined by electron microscopy measurements. Further heating, beyond 250 °C, induced thermal degradation of the polyurethane shell, and the remains at 350 °C weighed 10 ng. The insets of Figure 2b show optical microscopy images of the capsule captured in situ during the measurement. The size of the capsule decreased slowly with increasing temperature, because of the evaporation of chlorobenzene. The shape subsequently distorted as the rupture temperature approached, and the capsule shrank completely at temperatures above the degradation temperature of the shell. Figure 3 shows the variations in the weight percentage of the microcapsules as a function of temperature, measured using the commercial TGA and the microcantilevers. The microcantileverbased measurements showed a much sharper weight loss profile, with respect to temperature, than did the measurements using the commercial TGA. The sharpness of the change was attributed to the fact that the microcantilever measured the weight loss of a single capsule, whereas the commercial TGA measured only the average weight loss of a collection of capsules characterized by a range of rupture temperatures. Capsule-to-capsule rupture temperature variations may be induced by variations in capsule characteristics, such as the thickness and uniformity of the capsule shells, which is not easy to control during synthesis. No notable relation was found between rupture temperature and capsule size (see top portion of Figure S2 in the Supporting Information), indicating that capsules of the same size can have different thickness and uniformity of the capsule shells. Note that the weight percentage of the shell measured using the commercial TGA method was higher than that measured using the microcantilevers in Figure 3. The bundle sample used in the TGA measurement most likely included capsules that had broken during sample preparation, which generally induces the error in the determination of capsule composition. Because of this drawback, the microcantilever-based measurement of single-capsule properties provided a more accurate analysis of the capsule composition. Despite the variations in rupture temperature of each capsule, the thermal degradation temperature was determined to be constant across all capsules. Figure 4a shows the temperature-dependent frequency variations of the cantilever when two capsules were mounted. The diameters of the two capsules were ∼50 and 70 µm. The rupture temperature of each capsule was determined from the first derivative of the frequency, with respect to temperature. Figure 4b shows the mass change and the percentage weight loss Analytical Chemistry, Vol. 82, No. 13, July 1, 2010
5817
of chlorobenzene through the polyurethane shell at a given temperature is related to the permeability (P):23 ln
( )
mc(t) 3P )t mc(0) r
( )
(4)
where r is the initial capsule radius and mc(t) is the chlorobenzene mass at a given time t. The inset of Figure 5 shows variations in the normalized masses of the microcapsules, as a function of time. The permeabilities of capsules A, B, C, and D were calculated to be 3.7, 1.5, 0.66, and 0.49 nm/s, respectively, and the corresponding rupture temperatures were 88, 100, 113, and 132 °C, respectively, indicating that the rupture temperature decreases as the microcapsule permeability increases. However, the thicknesses of capsules A, B, C, and D, calculated using eq 3, were 1.15, 1.09, 1.06, 1.27 µm, respectively, which does not agree with the previous observation that microcapsules with low permeabilities typically have thick capsule shells.24 (Cross-sectional SEM images of broken capsules are shown in the bottom portion of Figure S2 in the Supporting Information.) This indicates that the permeabilities of individual capsules are affected not only by the shell thickness but also by the shell uniformity, as observed in the rupture temperatures of the capsules. Figure 4. (a) Temperature-dependent variations in the frequency (black trace) and the first derivative of the frequency (blue trace) of a cantilever on which two capsules were mounted. (b) Variations in the mass of the double-capsule system, as a function of temperature. The inset shows in situ optical microscopy images of the microcapsule-mounted cantilevers at 40, 80, 130, and 350 °C.
Figure 5. Temperature dependence of the weight percentage of the microcapsules using the microcantilevers; capsules A (black trace), B (red trace), C (green trace), and D (blue trace). The inset shows the variations in the normalized masses of chlorobenzene, as a function of time. Dotted line represents the temperature profile during heating.
calculated using eq 3. The weight loss of the larger capsule was three times greater than the weight loss of the smaller capsule at each rupture temperature. This loss ratio was comparable to the volume fraction of the core material for both capsules. Figure 5 shows the temperature dependence of the weight percentage of capsules with three different sizes: capsule A (73 µm), capsules B and C (62 µm), and capsule D (59 µm). In this experiment, the temperature was maintained at 70 °C for 20 min and then increased to 350 °C at a rate of 2 °C/min. The weight percent changes due to the evaporation of chlorobenzene at 70 °C were 23%, 12%, 7%, and 5% for capsules A, B, C, and D, respectively. The mass change that was due to the evaporation
5818
Analytical Chemistry, Vol. 82, No. 13, July 1, 2010
CONCLUSION In summary, we have demonstrated the first use of a silicon microcantilever for investigating the thermomechanical properties of single microcapsules. The mass changes of single microcapsules containing chlorobenzene were measured with unprecedented sensitivity, and the rupture temperature of the capsule and thermal degradation temperature of the polyurethane shell were clearly determined. Furthermore, the microcantilever-based system was capable of differentiating variations in the rupture temperatures of two capsules containing the same solvent. In addition, the permeability of the polyurethane shell, with respect to chlorobenzene, was measured, and the rupture temperature was observed to decrease with increasing permeability. This study can be extended to investigate the diffusion properties of a variety of solvents through polymer membranes as a function of temperature. ACKNOWLEDGMENT This research was supported by a grant (10162KFDA995) from Korea Food & Drug Administration in 2010. D.L. and Y.P. contributed equally to this work. SUPPORTING INFORMATION AVAILABLE Graph showing the variation in resonance frequency of a bare silicon microcantilever (Figure S1). Graph showing the variation in rupture temperature, relative to capsule size, and cross-sectional SEM images of broken capsules (Figure S2). This material is available free of charge via the Internet at http://pubs.acs.org. Received for review April 8, 2010. Accepted May 21, 2010. AC100913K (23) Richards, J. H. Polymer Permeability; Comyn, J., Ed.; Elsevier Applied Science Publishers Ltd.: New York, 1985; Chapter 6, p 235. (24) Yuan, L.; Liang, G.-Z.; Xie, J.-Q.; Li, L.; Guo, J. J. Mater. Sci. 2007, 42, 4390–4397.
