You are on page 1of 7

Journal of Colloid and Interface Science 263 (2003) 454460 www.elsevier.

com/locate/jcis

Inuence of solvent on the growth of ZnO nanoparticles


Zeshan Hu,a Gerko Oskam,b and Peter C. Searson a,
a Department of Materials Science and Engineering, Johns Hopkins University, Baltimore, MD 21218, USA b Departamento de Fsica Aplicada, CINVESTAV-IPN, Mrida, Yuc. 97310, Mexico

Received 7 June 2002; accepted 20 February 2003

Abstract We have synthesized ZnO nanoparticles by precipitation from zinc acetate in a series of n-alkanols from ethanol to 1-hexanol as a function of temperature. In this system, nucleation and growth are relatively fast and, at longer times, the average particle size continues to increase due to diffusion-limited coarsening. During coarsening, the particle volume increases linearly with time, in agreement with the Lifshitz SlyozovWagner (LSW) model. The coarsening rate increases with increasing temperature for all solvents and increases with alkanol chain length. We show that the rate constant for coarsening is determined by the solvent viscosity, surface energy, and the bulk solubility of ZnO in the solvent. 2003 Elsevier Inc. All rights reserved.
Keywords: ZnO; Nanoparticles; Nucleation; Growth; Coarsening; Ostwald ripening; Epitaxial attachment

1. Introduction The synthesis of ZnO quantum particles by precipitation from alcohols results in stable colloids of nanometer-sized particles [17]. In contrast, synthesis from aqueous solutions results in the formation of Zn(OH)2 [1,2] and hence ZnO can only be obtained by using stabilizers [1] or by subsequent hydrothermal treatment [8,9]. The nucleation process usually involves the reaction between a divalent zinc salt ZnX2 with hydroxide ions, where X represents the anion. In solution phase synthesis, processes such as coarsening and aggregation can compete with nucleation and growth in modifying the particle size distribution in the system [1012]. In previous work [6], we have shown that the nucleation and growth of ZnO from 2-propanol at room temperature is fast resulting in nanoparticles with an average radius of about 1.5 nm. After the supersaturation has been depleted and growth is completed, the average particle size continues to increase due to diffusion-limited coarsening [13,14]. The kinetics of nucleation and growth as well as processes such as coarsening and aggregation are expected to be strongly dependent on the properties of the solvent [1517]. Water is a dipolar, amphiprotic solvent with a high dielectric constant and, as a consequence, most salts are readily dissolved.
* Corresponding author.

Most alcohols are dipolar, amphiprotic solvents with a dielectric constant and viscosity that is dependent on the chain length [16]. In this paper, we report on the inuence of solvent on the synthesis of ZnO nanoparticles from zinc acetate at temperatures between 30 and 65 C. The inuence of the solvent provides a means to achieve control over the ZnO nanoparticle size and size distribution, which is essential for tailoring optical, electrical, chemical, and magnetic properties of nanoparticles for specic applications. 2. Materials and methods The ZnO colloids were prepared by precipitation from solution using Zn(CH3 CO2 )2 and NaOH. The overall reaction for the synthesis of ZnO nanoparticles from Zn(II) acetate can be written as Zn(CH3 CO2 )2 + 2NaOH ZnO + 2Na(CH3 CO2 )2 + H2 O. (1)

E-mail address: searson@jhu.edu (P.C. Searson). 0021-9797/03/$ see front matter 2003 Elsevier Inc. All rights reserved. doi:10.1016/S0021-9797(03)00205-4

The solvents used included ethanol (Pharmco, absolute ethanol), 1-propanol (J.T. Baker, reagent grade), 1-butanol (Alfa Aesar, reagent grade 99+%), 1-pentanol (Alfa Aesar, reagent grade 99+%), and 1-hexanol (Alfa Aesar, reagent grade 99+%). The solvents were used as received. The water contents of the solvents were determined by Karl

Z. Hu et al. / Journal of Colloid and Interface Science 263 (2003) 454460

455

Fisher titration (Brinkman 684 KF Coulometer): 25.2 mM (ethanol), 18.2 mM (1-propanol), 28.5 mM (1-butanol), 32.4 mM (1-pentanol), and 14.3 mM (1-hexanol). For a typical preparation, 1 mmol of zinc acetate dihydrate (Zn(CH3 CO2 )2 2H2 O; Aldrich, reagent grade) was dissolved in 80 ml of solvent in a covered ask under vigorous stirring at 50 C. After cooling to room temperature, 8 ml of the transparent zinc salt solution was added to 64 ml of the pure solvent. A 0.02 M NaOH (Aldrich, reagent grade) solution was prepared by adding sodium hydroxide (Aldrich, reagent grade) to the pure solvent in a covered ask under vigorous stirring at 60 C. After cooling to room temperature, 8 ml of the sodium hydroxide solution was added to 20 ml of the pure solvent. The covered asks containing the zinc acetate solution and the sodium hydroxide solution were heated to the growth temperature in a water bath. The sodium hydroxide solution was then added to the zinc acetate solution under vigorous stirring to give a total volume of 100 ml with 0.1 mmol of zinc acetate and 0.16 mmol of NaOH. From the overall reaction it follows that the synthesis is carried out with a 25% excess of Zn(II). Upon removal from the water bath, the colloids were covered and stored at room temperature. The colloids remained transparent and stable for periods of up to a few months, at which time they became translucent followed by the appearance of a ne white precipitate. Absorption spectra were obtained using a Shimadzu UV2101PC spectrophotometer. A slit width of 0.5 nm and a sampling interval of 0.2 nm were used to record the spectra between 275 and 400 nm. About 5-ml aliquots of the colloid were withdrawn during particle growth at predetermined time intervals and stored in an ice water bath prior to measurement. A blank solution of the solvent was used as reference. The solubility of ZnO was determined by atomic absorption spectroscopy (Perkin-Elmer Analyst 100). Samples were prepared as follows. ZnO powder (Aldrich, 99.9%) with particle sizes smaller than 1 m was suspended in 2-propanol and ultrasonically agitated for 1 h. The suspension was then ltered over a 0.1-m lter (Gelman Sciences) and the remaining solid was collected. The resulting ZnO particles with sizes larger than 0.1 m were suspended in 2-propanol for 3 days at room temperature. The suspension was ltered over a 0.1-m lter before analysis of the Zn(II) concentration with atomic absorption spectroscopy (AAS). The multiple ltration steps are necessary since nanoparticles are also detected by AAS, which would result in an overestimation of the ZnO solubility. Calibration was performed by preparing solutions of Zn(ClO4 )2 of known concentration in the mixed solvents. 3. Results and discussion Figure 1 shows absorption spectra for the synthesis of ZnO from zinc acetate in a series of alcohols from ethanol to 1-hexanol at 35 C. In general, the absorbance spectra

Fig. 1. Absorbance spectra for ZnO colloids at 35 C as a function of time (from left to right): ethanol (30, 60, 90, 120 min); 1-propanol (2, 4, 8, 16, 30, 60, 93, 120 min); 1-butanol (2, 4, 8, 16, 30, 60, 90, 120 min); 1-pentanol (2, 4, 8, 16, 30, 67, 95, 120 min); 1-hexanol (2, 4, 8, 16, 30, 75, 95, 122 min).

exhibit a well-dened exciton peak and absorbance onset characteristic of solid ZnO. In all cases, the absorption onset is signicantly blue-shifted compared to the absorption onset for bulk zinc oxide at about 385 nm, showing that the average particle size is in the quantum regime. For all solvents, the absorption onset red-shifts with time due to the increase in average particle size. The inuence of the solvent can be clearly seen during the nucleation process after mixing the zinc acetate and sodium hydroxide solutions. For the longer chain length alcohols, the spectra show a well-dened absorbance onset immediately after mixing, indicating that nucleation is fast. In contrast, for the shorter chain length alcohols, ethanol and 1-propanol, the absorbance spectra evolve with time, indi-

456

Z. Hu et al. / Journal of Colloid and Interface Science 263 (2003) 454460

Fig. 3. Transmission electron microscope image of ZnO particles grown from Zn(CH3 COO)2 in propanol at 55 C for 8.5 h.

Fig. 2. Band gap, E , and absorption onset wavelength, , for ZnO quantum particles versus particle radius, calculated from Eq. (1), using me = 0.26, mh = 0.59, and = 8.5. The onset wavelength was calculated from E = hc/.

cating that nucleation and growth are slower. The acceleration of the nucleation and growth with decreasing dielectric constant has been reported previously for ZrO2 particles in mixed aqueous/alcohol solution [18,19]. The average particle size in a colloid can be obtained from the absorption onset using the effective mass model (e.g., [20]) where the band gap E (in eV) can be approximated by
bulk E = Eg +

1 h 1 2 2 + 2 2er me m0 mh m0

1 .8 e 40r , (2)

1 0.124e3 1 + 2 2 h (40 ) me m0 mh m0

bulk is the bulk band gap (eV), h is Planks constant, where Eg r is the particle radius, me is the electron effective mass, mh is the hole effective mass, m0 is free electron mass, e is the charge on the electron, is the relative permittivity, and 0 is the permittivity of free space. Figure 2 shows the band gap and corresponding absorption onset wavelength versus particle radius calculated using Eq. (2). Due to the relatively small effective masses for ZnO (me = 0.26, mh = 0.59 [21,22]), band gap enlargement is expected for particle radii less than about 4 nm. Figure 3 shows a transmission electron microscope image of ZnO particles grown from Zn(CH3 COO)2 at 55 C after 8.5 h. From the lattice fringes it is seen that the particles are approximately spherical and that many particles exhibit some faceting. From analysis of many images we de-

termine an average diameter of 6.5 1.2 nm (289 particles). The average particle diameter obtained from the absorption onset is 6.9 nm, in good agreement with the value obtained from analysis of transmission electron microscope images. We note, however, that the particle diameter is in the range where the band gap enlargement is small and hence the error in determining the particle size is relatively large. Figure 4 shows the time dependence of the average particle radius, obtained from the absorption onset and Eq. (2), at different temperatures. In all solvents, the average radius increases with time. In addition, at any time it is seen that the particle size increases with increasing temperature. For all solvents the initial particle radius is between 1.2 and 1.5 nm depending on temperature, indicating that the mechanism for nucleation is similar in all cases. Coarsening involves the growth of larger crystals at the expense of smaller crystals and is governed by capillary effects. Since the chemical potential of a particle increases with decreasing particle size, the equilibrium solute concentration for a small particle is much higher than that for a large particle. The resulting concentration gradients lead to the transport of solute (e.g., metal ions) from the small particles to the larger particles. The rate law for this process, derived by Lifshitz, Slyozov, and Wagner (LSW) [23,24], is given by
3 0 = kt, r 3 r

(3)

where r is the average particle radius, r 0 is the average initial particle radius, k is the rate constant, and t is time. Figure 5 shows the growth data from Fig. 3 re-plotted as r 3 versus time. The linear region at longer times is consistent with rate law for coarsening and indicates that particle growth is completed shortly after nucleation so that at longer times the increase in particle size is determined solely by diffusionlimited coarsening. Since the slope of the linear region of the curves corresponds to the rate constant for coarsening, it is evident that the rate constant is dependent on the temperature and solvent.

Z. Hu et al. / Journal of Colloid and Interface Science 263 (2003) 454460

457

Fig. 4. ZnO particle radius versus time determined from the results in Fig. 1 and Eq. (1) for different temperatures: (4) 75 C, (!) 65 C, (1) 55 C, (e) 45 C, (P) 35 C, (E) 21 C, and (%) 0 C.

Fig. 5. The results from Fig. 3 re-plotted as r 3 versus time at (4) 75 C, (!) 65 C, (1) 55 C, (e) 45 C, (P) 35 C, (E) 21 C, and (%) 0 C.

Figure 6 shows the rate constant k versus 1/T for growth from the series of alcohols determined from Fig. 5. The rate constant is strongly temperature dependent and increases with increasing alcohol chain length. The rate constant k is given by [6] k=
2c 8 DVm r = , (4) 9RT where is the surface energy, D is the diffusion coefcient, Vm is the molar volume, and cr = is the equilibrium concentration at a at surface (i.e., the bulk solubility). The

temperature and solvent dependence of the rate constant can be considered by eliminating the diffusion coefcient from Eq. (4) using the StokesEinstein equation, D= kB T , 6a (5)

where kB is the Boltzmann constant, is the viscosity of the solvent, and a is the solvated ion radius. Substituting into Eq. (4), we obtain k=
2c 8 Vm r = . 54aNA

(6)

458

Z. Hu et al. / Journal of Colloid and Interface Science 263 (2003) 454460

Fig. 7. Viscosity versus temperature for (a) ethanol, (b) 1-propanol, (c) 1-butanol, (d) 1-pentanol, and (e) 1-hexanol. Data obtained from Ref. [33]. Fig. 6. The rate constant for coarsening determined from the slopes of the r 3 versus time curves versus inverse temperature for (%) ethanol, (1) 1-propanol, (P) 1-butanol, (!) 1-pentanol, and (e) 1-hexanol.

Thus, the rate constant for coarsening is a function of (the surface energy of the crystal), Vm (the molar volume), cr = (the bulk solubility), (the solvent viscosity), and a (the solvated ion radius). The molar volume for ZnO is 14.8 cm3 mol1 and is essentially constant since the linear expansion coefcient is about 4 106 K1 over a wide temperature range [22]. The surface energy for the solid vapor interface for metal oxides is on the order of 1 J m2 [2527]; however, adsorption of ions and solvent molecules is expected to reduce the surface energy of the solidliquid interface to values in the range 0.10.5 J m2 [28]. In the gas phase, methanol, ethanol, and propanol adsorb dissociatively on ZnO under ambient conditions [2931]; however, data for the surface energy of solids in alcohols are not available in the literature. The solvated ion radius for Zn(II) in methanol is 0.51 nm and is independent of temperature over the temperature range of interest [32]. The ionic radius for Zn(II) is 0.075 nm and the length of a methanol molecule is about 0.36 nm. Thus, the solvated ion radius implies a high coordination number of the solvent molecules. Since 1-hexanol is about twice as long as an ethanol molecule, the solvated ion radius for zinc ions in hexanol would be expected to be about twice the solvated ion radius in ethanol, assuming similar coordination numbers. From Eq. (6) we see that k a 1 , so that an increase in the solvated ion radius due to increasing chain length would be expected to result in a decrease in the rate constant by a factor of about 2. However, from Fig. 6 it can be seen that the rate constant increases with increasing chain length, indicating that the inuence of the solvent on the growth kinetics is not dominated by the solvated ion radius. Hence, the dependence of the rate constant on the chain length of the solvent must be dominated by the surface energy , the solvent viscosity , or the bulk solubility cr = . Figure 7 shows that the viscosity of the alcohols is strongly dependent on temperature and chain length, especially at lower temperatures [33]. Figure 8 shows the rate constant, k , plotted versus the solvent viscosity. At each tem-

Fig. 8. The rate constant for coarsening versus the viscosity of the solvent: (%) ethanol, (1) 1-propanol, (P) 1-butanol, (!) 1-pentanol, and (e) 1-hexanol. The solid lines connect the data points for the same solvent, while the dotted lines connect the data points for constant temperature.

perature, the rate constant increases with solvent viscosity (solvent chain length). From Eq. (6) it is apparent that the rate constant k 1 if cr = and are independent of the solvent. This dependence is related to the fact that the diffusion coefcient for zinc ions in the solvent decreases with increasing viscosity. However, Fig. 8 shows that the rate constant, at constant temperature, increases with increasing viscosity according to a power law relationship with exponents increasing from 0.21 at 0 C to 0.79 at 75 C. The positive exponents in Fig. 8 imply that either the surface energy or the bulk solubility cr = must increase with increasing viscosity (solvent chain length). For each solvent, the rate constant decreases with increasing viscosity (decreasing temperature) with an exponent that decreases from 2.1 for ethanol to 1.33 for hexanol. For a given solvent, an exponent of 1 is predicted from Eq. (6) if cr = is independent of temperature (Vm , , and a are expected to be weakly dependent on temperature). Thus, the magnitude of the exponents indicate that, for a given solvent, cr = increases with decreasing viscosity (increasing temperature), as would be expected for most solids. The bulk solubility of ZnO is expected to be related to the enthalpy of solution, which contains contributions from

Z. Hu et al. / Journal of Colloid and Interface Science 263 (2003) 454460

459

Fig. 9. The solubility of ZnO in mixtures of 2-propanol and water determined versus inverse dielectric constant of the solvent mixture.

the lattice enthalpy and the ion solvation enthalpy. Since the lattice enthalpy of ZnO is independent of solvent, the inuence of the solvent on the bulk solubility is expected to be through the enthalpy of solution. The bulk solubility cr = can be estimated from the coarsening rate constants (Fig. 6) and Eq. (6). Taking = 0.1 J m2 , Vm = 14.8 cm3 mol1 , a = 0.51 nm, = 0.010.1 P, and k = 104 103 nm3 s1 , the bulk solubility is determined to be in the range 1011 109 mol l1 , and as a consequence cannot be measured directly. To determine the dependence of the solubility of ZnO on the dielectric constant of the solvent, we measured the solubility of ZnO in water/2-propanol mixtures as a function of the concentration of water. These equilibrium solubility experiments were performed in solvent mixtures, thus minimizing the inuence of ion adsorption. Figure 9 shows the zinc ion concentration versus water concentration in water/2-propanol mixtures. Since the dielectric constant increases with increasing concentration of water, Fig. 9 shows that the solubility of the ZnO particles increases with increasing dielectric constant . This dependence indicates that the partially covalent nature of the bonding in ZnO plays an important role in determining the dissolution mechanism since, for a purely ionic solid, the solubility would be expected to increase with increasing solvent dielectric constant. In pure water, the measured Zn(II) concentration is 49 M, in good agreement with values reported in the literature [34]. For water concentrations less than 60 vol%, the Zn(II) concentration is below the detection limit for AAS (1 M). Estimation of the solubility of Zn(II) in pure propanol by extrapolation of the solubility data to zero water content results in values several orders of magnitude lower than that in pure water [35], consistent with the values of 1011 109 mol l1 obtained from the rate constant using Eq. (6). Figure 10 shows the growth rate constant multiplied by the solvent viscosity, k, versus 1 . It can be seen that k increases with increasing 1 in a similar manner for all alcohols. At low 1 (i.e., high ) k increases strongly with 1 , while at higher 1 (i.e., lower ) k is essentially independent of 1 . Figure 10 shows that the k versus 1 curves for all solvents and temperatures collapse onto a uni-

Fig. 10. The rate constant for coarsening multiplied with the solvent viscosity for ZnO particle growth versus inverse dielectric constant for (%) ethanol, (1) 1-propanol, (P) 1-butanol, (!) 1-pentanol, and (e) 1-hexanol.

versal curve except at high viscosity. These results illustrate that the rate constant multiplied by the solvent viscosity is uniquely described as a function of the dielectric constant. From Eq. (6) it can be seen that k is proportional to the product cr = assuming that Vm and a are constant. According to Eq. (6), the increase of k with inverse dielectric constant implies that cr = increases with decreasing . Since cr = is expected to increase with increasing dielectric constant, as described above, the results in Fig. 10 suggest that the increase in k with increasing 1 is dominated by an increase in surface energy with increasing (i.e., the surface energy increases with increasing chain length of the solvent). These results show that particle growth and coarsening are strongly dependent on solvent through the viscosity, bulk solubility, and surface energy. The solvent is an important parameter in determining the coarsening kinetics, and that by choice of solvent, the ZnO particle synthesis can be tuned.

4. Summary The synthesis of ZnO quantum particles by precipitation from a series of n-alkanols from ethanol to 1-hexanol alcohols results in stable colloids of nanometer-sized particles. For ethanol and 1-propanol, nucleation and growth are retarded compared to longer chain length alcohols where nucleation and growth are fast. After the supersaturation has been depleted and nucleation and growth are completed, the average particle size continues to increase due to diffusionlimited coarsening. For all alcohols, the kinetics of coarsening were consistent with the LifshitzSlyozovWagner model. The coarsening rate constant increases with increasing temperature and longer solvent chain length, due to the inuence of solvent viscosity, surface energy, and the bulk solubility of ZnO. These results illustrate that the solvent is an important parameter in controlling particle size.

460

Z. Hu et al. / Journal of Colloid and Interface Science 263 (2003) 454460

Acknowledgments The authors gratefully acknowledge support from the JHU MRSEC (NSF Grant No. DMR00-80031). G.O. acknowledges support from CONACYT (Grant No. 33171-E (G.O.)).

References
[1] U. Koch, A. Fojtik, H. Weller, A. Henglein, Chem. Phys. Lett. 122 (1985) 507. [2] D.W. Bahnemann, C. Kormann, R. Hoffmann, J. Phys. Chem. 91 (1987) 3789. [3] M. Haase, H. Weller, A. Henglein, J. Phys. Chem. 92 (1988) 482. [4] L. Spanhel, M.A. Anderson, J. Am. Chem. Soc. 113 (1991) 2826. [5] P.V. Kamat, B. Patrick, J. Phys. Chem. 96 (1992) 6829. [6] E.M. Wong, J.E. Bonevich, P.C. Searson, J. Phys. Chem. B 102 (1998) 7770. [7] L. Guo, S. Yang, C. Yang, P. Yu, J. Wang, W. Ge, G.K.L. Wong, Appl. Phys. Lett. 76 (2000) 2901. [8] C.-H. Lu, C.-H. Yeh, Ceram. Int. 26 (2000) 351. [9] D. Chen, X. Jiao, G. Cheng, Solid State Commun. 113 (2000) 363. [10] V.K. LaMer, R.H. Dinegar, J. Am. Chem. Soc. 72 (1950) 4847. [11] J. Park, V. Privman, E. Matijevic, J. Phys. Chem. B 105 (2001) 11630. [12] R.L. Penn, G. Oskam, T.J. Strathmann, P.C. Searson, A.T. Stone, D.R. Veblen, J. Phys. Chem. B 105 (2001) 2177. [13] Z. Hu, G. Oskam, R.L. Penn, N. Pesika, P.C. Searson, J. Phys. Chem. B, submitted. [14] G. Oskam, Z. Hu, R.L. Penn, P.C. Searson, Phys. Rev. E, submitted. [15] O. Popovich, R.P.T. Tomkins, Nonaqueous Solution Chemistry, Wiley, New York, 1981. [16] C. Kalidas, G. Hefter, Y. Marcus, Chem. Rev. 100 (2000) 819.

[17] J. Barthel, H.-J. Gores, in: G. Mamantov, A.I. Popov (Eds.), Chemistry of Nonaqueous Solutions, VCH, New York, 1994, pp. 1112. [18] M.Z.-C. Hu, M.T. Harris, C.H. Byers, J. Colloid Interface Sci. 198 (1998) 87. [19] M.Z.-C. Hu, E.A. Payzant, C.H. Byers, J. Colloid Interface Sci. 222 (2000) 20. [20] L.E. Brus, J. Phys. Chem. 90 (1986) 2555. [21] S. Shionoya, in: S. Shionoya, W.M. Yen (Eds.), Phosphor Handbook, CRC, Boca Raton, FL, 1998. [22] L.I. Berger, Semiconductor Materials, CRC, Boca Raton, FL, 1997. [23] I.M. Lifshitz, V.V. Slyozov, J. Phys. Chem. Solids 19 (1961) 35. [24] C. Wagner, Z. Elektrochem. 65 (1961) 581. [25] X.G. Wang, W. Weiss, Sh.K. Shaikhutdinov, M. Riter, M. Petersen, F. Wagner, R. Schlogl, M. Schefer, Phys. Rev. Lett. 81 (1998) 1038. [26] I. Manassidis, A. De Vita, M.J. Gillan, Surf. Sci. Lett. 285 (1993) L517. [27] D.A. Weirauch, P.D. Ownby, J. Adhesion Sci. Technol. 13 (1999) 1321. [28] A. Zangwill, Physics at Surfaces, Cambridge University Press, Cambridge, 1988. [29] J.M. Vohs, M.A. Barteau, Surf. Sci. 201 (1988) 481. [30] J.M. Vohs, M.A. Barteau, Surf. Sci. 221 (1989) 590. [31] J.M. Vohs, M.A. Barteau, J. Phys. Chem. 95 (1991) 297.29. [32] R. Lovas, G. Macri, S. Petrucci, J. Am. Chem. Soc. 92 (1970) 6502. [33] J. Barthel, R. Neueder, R. Meier (Eds.), Electrolyte Data Collection, Part 3: Viscosity of Nonaqueous Solutions I: Alcohol Solutions, Chemistry Data Series, Vol. XII, Dechema, Frankfurt, 1997, Part 3. [34] H. Remy, Z. Elektrochem. Angew. Phys. Chem. 31 (1925) 88. [35] Extrapolation of an exponential t to the experimental data gives a solubility for Zn(II) in pure propanol of 3.5 107 M. From Fig. 10 it is clear that the solubility decreases more rapidly than predicted by the exponential t so that the extrapolation represents an upper limit. Since there are no suitable models that we can use to t the data, we can only conclude that these data are consistent with the solubilities determined from the rate constants.

You might also like