Temperature Dependence of the Dielectric ... - ACS Publications


Temperature Dependence of the Dielectric...

4 downloads 99 Views 1MB Size

Ind. Eng. Chem. Res. 2010, 49, 3011–3014

3011

Temperature Dependence of the Dielectric Properties of 2,2′-Azobis(2-methyl-butyronitrile) (AMBN) Alastair D. Smith,† Edward H. Lester,† Kristofer J. Thurecht,‡,§ Sam W. Kingman,† Jaouad El Harfi,†,‡ Georgios Dimitrakis,† John P. Robinson,*,† and Derek J. Irvine*,†,‡ National Centre for Industrial MicrowaVe Processing, Faculty of Engineering, Department of Chemical and EnVironmental Engineering, UniVersity of Nottingham, Nottingham, NG7 2RD, U.K., and School of Chemistry, UniVersity of Nottingham, UniVersity Park, Nottingham, NG7 2RD, U.K.

For the first time, the response of 2,2′-azobis(2-methyl-butyronitrile) (AMBN) to microwave heating at 2.45 GHz was studied by measuring its dielectric properties as a function of temperature with the observed variations being attributed to the physical form of the initiator, the 1 h decomposition half-life temperature, and the relaxation frequency. 1. Introduction AMBN is an initiator commonly used in both laboratory and commercial processes to conduct free radical polymerization. This paper reports the first measurements of the dielectric properties of a free radical polymerization (FRP) initiator conducted in the bulk. When FRP is conducted using conventional heating, it is important to fully understand the mechanism and time scales for the initiator to decompose and produce the initial radical species. This ensures that a safe, efficient, and well-controlled polymerization results. Likewise, it is equally important to identify how this decomposition is influenced by microwave energy, if this is the heating source to be employed. Only once the correct decomposition characteristics have been identified can any specific or “unique” effect which is due to the use of microwaves be identified and fully exploited. Studies in the field of microwave-assisted polymerization have received an increasing level of study in recent years. Literature reports have included studies in the areas of step-growth,1 ring-opening,2 and free radical polymerization.3 This has led to several comprehensive literature reviews.4,5 Industrial FRP predominantly uses thermal initiators to conduct large scale commercial polymerizations. The initiators utilized industrially contain a thermally sensitive, “fragile” bond which undergoes homolytic bond scission within a particular temperature range. This range is defined by the chemical structure of the initiator. In commercial processes, initiator performance is typically ranked by the 1 h half-life temperature (t1/2 ) 1 h), which is a reference to the temperature at which half the initiator decomposes within 1 h. In this study, we investigate the dielectric properties of 2,2′azobis(2-methyl-butyronitrile) (AMBN), a widely used vazo initiator over a temperature range of 25-150 °C. This encompasses the typical reaction temperatures used for FRP reactions with this initiator as t1/2 ) 1 h is 84 °C.6 The structure of AMBN is shown in Figure 1. This study has been carried out to establish the nature of any differences between the decomposition of * To whom correspondence should be addressed. Tel.: 00441159514088 (D.J.I.), 00441159514092 (J.P.R.). Fax: 00441159514115 (D.J.I.), 00441159514115 (J.P.R.). E-mail: [email protected] (D.J.I.), [email protected] (J.P.R.). † National Centre for Industrial Microwave Processing. ‡ School of Chemistry. § New Address: Australian Institute for Bioengineering and Nanotechnology and Centre for Magnetic Resonance, The University of Queensland, QLD, 4072, Australia.

AMBN under microwave and conventional heating. This has been achieved by measuring how the molecule interacts with an applied electromagnetic field at a set frequency across the stated range of temperatures. The AMBN used in this study was purchased from Wako, and the purity and absence of entrained water was confirmed by nuclear magnetic resonance (NMR) techniques. The response of a material to electromagnetic energy can be quantified by the dielectric properties of that material. The dielectric properties are related to the complex permittivity, ε*, which is an expression containing both real and imaginary parts. The real part is known as the dielectric constant, ε′, and is related to the ability of a material to be polarized or store electromagnetic energy. The imaginary part is termed the dielectric loss factor, ε′′, and expresses the ability of that material to convert stored electromagnetic energy to heat. The loss tangent (tan δ) can be used to quantify the extent to which a material absorbs microwave energy. Systems that contain a mixture of microwave transparent and microwave absorbent components will undergo selective heating in a microwave field, with those components with high ε′′ being initially heated selectively over lower loss components. However in attempting to identify microwave effects, care must be taken to ensure that heat transfer from the selectively heated components to those in the mixture that are less absorbing does not occur, least it masks/suppresses the microwave inspired effect. Thus the dielectric properties of a specific compound will vary with frequency and temperature, so it is vital to understand the variation in dielectric properties across a broad range of conditions to fully characterize the behavior of that compound in an applied electromagnetic field. Similarly, in a chemical reaction, the physical state, chemical composition, and temperature of the system can change, so it is imperative that the dielectric properties of each of the system components are measured at temperatures which are representative of the reaction conditions, so that the true influence of microwaves can be understood. This study of pure AMBN represents the first steps to characterize the influence of microwave energy on FRP reactions.

Figure 1. Structure of AMBN.

10.1021/ie901389a  2010 American Chemical Society Published on Web 02/23/2010

3012

Ind. Eng. Chem. Res., Vol. 49, No. 6, 2010

Figure 2. Schematic of the cavity perturbation experimental apparatus.

2. Experimental Method The dielectric properties of AMBN were measured using a cavity perturbation technique. The technique works by measuring the frequency shift and the change in quality factor when the test material is introduced into a cylindrical cavity which is resonating at a particular frequency, in this case 2.45 GHz (see Figure 3). For the cavity dimensions used in the experiments reported in this paper, this frequency corresponds to the TM050 mode. The sample is introduced into the center of the cavity were the electric field attains the maximum value, this ensures maximum interaction between the sample and the field. Figure 2 shows a schematic of the cavity as operated. The sample is inserted into a quartz tube of 3 mm internal diameter. The mass of the sample is measured using a fourfigure balance, and its volume is calculated from a height measurement using a traveling microscope. A furnace is located above the cavity, in which the sample resides until it has equilibrated at the required test temperature. At this point, a step-motor is used to rapidly move the sample into the cavity, where its presence causes a change in the resonant frequency compared to the empty cavity. A vector network analyzer is used to detect the frequency shift, δf, and the resultant quality factor, Q, and the dielectric constant and dielectric loss factor and loss tangent can then be calculated using eqs 1-3.7-9 ε′ ) 1 + 2J12(x1,m) ε′′ ) J12(x1,m)

(

δf Vc fo Vs

(1)

)

(2)

1 1 Vc Q1 Q0 Vs

tan δ )

ε′′ ε′

(3)

Where J1(x1,m) is the second order root of the first kind Bessel function, fo is the resonant frequency, Vc is the volume of the resonant cavity, and Vs is the sample volume. The cavity perturbation technique is best suited for measuring low complex permittivity materials such as PTFE and in this case gives a relative percentage error in the measurement of less than 1%. The relative percentage error is calculated by the formula shown in eq 4. error )

|V - Vmeas | × 100 V

(4)

Where V is the actual value and Vmeas is the measured value. However, for high complex permittivity materials such as ethanol (ε′ ) 6.63, ε′′ ) 6.43 at 2.45 GHz and 20 °C), the relative percentage errors in the determination of complex

permittivity can increase to 16.48% and 23.37% for the real and the imaginary part, respectively (for a sample 3 mm in diameter and 10 mm in height). Thus, the physical dimensions and the overall volume of the sample can significantly influence the measurement precision, and this is particularly true for materials of higher complex permittivity. This problem can be remedied by adjusting the volume of the sample accordingly (mainly reducing the diameter of the sample) and by ensuring that the diameter to length ratio of the sample is in region of 0.2-0.3. The measurement process is repeated by removing the sample from the cavity back to the furnace until the next required temperature is achieved, before being reintroduced to the cavity and collecting the measurement at the new temperature. The total time that the sample spends out of the furnace and in the cavity during this procedure is of the order of 3 s. Therefore, heat loss is considered to be negligible. In practice, the average time between sequential measurements is approximately 15 min. 3. Results and Discussion The dielectric constant and dielectric loss factor of AMBN at temperatures from 25-150 °C are shown in Figures 4 and 5, respectively. In both cases, the volume change in the sample during the temperature sweep was quantified and the data was corrected using eqs 1 and 2. The results reported in this communication are an average of five different experiments. The maxima and minima are also reported alongside the average values. The dielectric properties of AMBN depend strongly upon temperature, and the trends are highly nonlinear. At temperatures up to 35 °C, the dielectric loss factor is very low; however, both the dielectric constant and loss factor undergo a sharp increase around 45 °C. Between 45 and 100 °C, the dielectric constant steadily increases, whereas the loss factor initially decreases. At about 75 °C, the loss factor increases up to around 100 °C, at which point the loss factor undergoes a steady decrease. The trends shown in Figures 4 and 5 result from the influence of a number of competing physical and chemical effects, which are explained below. Figure 6 shows a plot of tan δ with temperature for AMBN, along with the change in sample mass over the same temperature range. The dielectric property data shown in Figure 5 can be split into a number of distinct regions: 3.1. Temperatures up to 45 °C. The melting point of AMBN is 49-51 °C.6 In solid form AMBN is essentially transparent to microwaves, as evidenced by the very low tan δ values below the melting point. Once 45 °C is reached, the tan δ value increases sharply, while there is no change in sample mass at this point. AMBN contains CtN groups, and when subjected to an alternating electric field, the dipoles of the CtN groups will attempt to align with the field, causing intra- and intermolecular motion (i.e., movement relative to nearest neighbors) of the molecule. However, in solid form, this molecular movement is restricted due to the presence of relatively strong intermolecular forces and hence the dipoles are unable to significantly align with the very fast alternating electric field at the microwave frequencies, thus the low tan δ value. 3.2. Temperatures from 45 to 80 °C. Once the melting point is reached and the AMBN is liquid, the restrictions on its intermolecular motion are reduced. In liquid form, the dipoles on the CtN group move to align themselves with the alternating electric field, causing molecular rotation. However, due to the presence of forces exerted by adjacent molecules, the attempted

Ind. Eng. Chem. Res., Vol. 49, No. 6, 2010

3013

Figure 3. Electric and magnetic fields of the resonating TM050 mode at a frequency of 2.45 GHz.

Figure 4. Plot of the dielectric constant of AMBN at 2.45 GHz against temperature.

Figure 5. Plot of the dielectric loss factor of AMBN at 2.45 GHz against temperature.

alignment with the electric field is not perfectly in phase and this gives rise to heat losses due to intermolecular friction within the material. Therefore, in liquid form AMBN absorbs electromagnetic energy, and converts it into heat. This is evidenced by the correlation of the first step-change in tan δ in Figure 4 with the reported melting point for AMBN (within ∼4 °C). This slight difference being an artifact of the temperature control system used on the furnace. Similar behavior is observed with H2O, where at 2.45 GHz liquid water (ε′′ ) 13) is a much stronger absorber of microwave energy than ice (ε′′ ) 0.02).10 Of note is the behavior between 45 and 80 °C. Figure 4 shows that the dielectric constant increases with temperature in this region, whereas Figure 5 shows a corresponding decrease in loss factor. Physically, the decrease in dielectric loss factor could be due to a decrease in the resistance to molecular movement

Figure 6. Effect of temperature on tan δ (2.45 GHz) and mass change of AMBN: (s) mp; (- - -) t1/2 ) 1 h.

as the viscosity of the system reduces with increasing temperature, meaning that less energy is dissipated as heat. Alternatively, the observations in Figure 5 are consistent with the Debye relaxation model, where the relaxation time of the system decreases with increasing temperature provided that the frequency in which the measurement is carried out is well below the relaxation frequency of the system.11 One key prediction that has been made from comparing the thermogramitmetric analysis (TGA) and dielectric property data in the room temperature to 80 °C region is that exposure to microwaves at this frequency will not lead to additional decomposition of the AMBN. The decomposition of AMBN is governed by the conventional heating. Previous studies12,13 have shown differential behavior of free radical initiators in microwave experiments, and thus, it could be concluded that rapid microwave induced initiator decomposition explained the empirical, synthesis based observations made. The measurements reported in this paper indicate that this is not the case. 3.3. 80-100 °C. Above 80 °C, the dielectric loss factor increases further. However, in this region, a significant decrease in the sample mass is also observed (Figure 6). Across this temperature range, the AMBN decomposes to form free radicals at an appreciable rate, the 1 h half-life temperature being 84 °C.6 In this decomposition process, 1 mol of N2 and 2 mol of radicals are evolved for every mole of AMBN that decomposes. Thus stoichiometrically, a 16.7% decrease in mass would be expected due to this evolution of N2, and this corresponds with the data shown in Figure 6. Additionally, the free radical species produced now possess an unpaired electron which introduces an additional dipole to that of the CtN group. This will also respond to the alternating electric field, causing the molecule to rotate and dissipate energy as heat. The combination of the

3014

Ind. Eng. Chem. Res., Vol. 49, No. 6, 2010

free radical and the CtN group means that the decomposition product has a higher dielectric loss factor than the liquid AMBN present in the sample at lower temperatures. 3.4. Above 100 °C. At temperatures >100 °C, the half-life of AMBN decreases significantly (t1/2 ) 0.1 h at 104 °C).6 This increase in radical generation is expected to lead to a more rapid decay of radicals via radical-radical combination. After decay, the unpaired electron is no longer present having been used to form a new covalent bond. Therefore the polarity of the molecule is significantly reduced, and thus, the dielectric loss factor decreases. The decay products still contain the CtN groups; hence, the presence of the remaining dipoles means that the decay product is microwave absorbent to a similar degree as the liquid AMBN. Thus the dielectric properties recorded in this region will be a combination of the materials present including the decay products. Addtionally, the continuing mass loss at higher temperatures is likely to be due to evaporative loss of the decay products rather than further decomposition. This mass loss of sample will also influence the dielectric property measurement results. To ensure that the dielectric property results were not being adversely effected by the time that it took to conduct the temperature ramp experiment fresh samples of vazo initiator were heated directly to 80 and 100 °C to reduce the time taken to reach the target temperature. The dielectric response for the 80 °C sample was not found to be significantly different from the ramp experiment. However there was a statistically significant difference between the ramp and direct results at 100 °C. This is thought to be due the time that the sample spends in the oven relative to it is half-life at that temperature. Therefore, because the initiator spends less time at these elevated temperatures when heated directly this means that (a) there is a greater level of live radical still present in the directly heated mixture thus giving a different dielectric response than the ramp experiment where the mixture has more decay products present and (b) less evaporative loss will have occurred. However direct heating only accentuated the increase in the level of response observed between 80 and 100 °C; thus, all the conclusions made from the ramp experiment remain valid. 4. Conclusions This study is the first to clearly demonstrate that the dielectric properties of AMBN vary significantly with temperature and in a highly nonlinear fashion. Furthermore, it is clear that microwave energy does not encourage rapid decomposition of the initiator at lower temperatures than would normally be expected, which is a conclusion that could have been drawn from the empirical studies in earlier publications. Rather the decomposition of AMBN occurs due to conventional heating. The AMBN is essentially microwave transparent when solid and will only absorb microwave energy to an appreciable degree

when in liquid form. AMBN is most receptive when the free radicals are formed and absorbs less once the radicals start to decay. Therefore it is clear from this study that the design or selection of a microwave system for accelerating FRP reactions must be based upon the dielectric properties of the entire system and, therefore, knowledge of their behavior with temperature is of paramount importance. Acknowledgment We acknowledge the following for funding: the EPSRC for a studentship (A.D.S.), the DICE initiative (D.J.I., J.E.H.), and the Royal Commission for the Exhibition of 1851 for a research fellowship (K.J.T.). Literature Cited (1) Hoogenboom, R.; Schubert, U. S. Microwave-Assisted Polymer Synthesis: Recent Developments in a Rapidly Expanding Field of Research. Macromol. Rapid Commun. 2007, 28, 368. (2) Odian, G. Principles of Polymerization, 4th ed.; Wiley: New York, 2004. (3) Jovanovic, J.; Adnadjevic, B. Comparison of the Kinetics of Conventional and Microwave Methyl Methacrylate Polymerization. J. Appl. Polym. Sci. 2007, 104, 1775. (4) Zhang, C.; Liao, L.; Gong, S. Recent Developments in MicrowaveAssisted Polymerization with a Focus on Ring-Opening Polymerization. Green Chem. 2007, 9, 303. (5) Wiesbrock, F.; Hoogenboom, R.; Schubert, U. S. Microwave-Assisted Polymer Synthesis: State-of-the-Art and Future Perspectives. Macromol. Rapid Commun. 2004, 25, 1739. (6) Brandrup, J.; Immergut, E. H. Polymer Handbook, 4th ed.; WileyInterscience: New York, 1998. (7) Horner, F.; Taylor, T. A.; Dunsmuir, R.; Lamb, J.; Jackson, W. Resonance Methods of Dielectric Measurement at Centimetre Wavelength. J. Inst. Elec. Eng. 1946, 53. (8) Damaskos, N. J.; Kelsall, B. J. Measuring Dielectric Constants of Low-Loss Materials using a broadband cavity tecnique. MicrowaVe J. 1995, 140. (9) Li, S.; Akyel, C.; Bosisio, R. G. Precise Calculations and Measurements on the Complex Dielectric Constant of Lossy Materials Using TM010 Cavity Perturbation Techniques. IEEE Trans. MicrowaVe Theory Tech. 1981, 29, 1041. (10) Meredith, R. The Engineers’ Handbook of Industrial MicrowaVe Heating, 1st ed.; The Institution of Electrical Engineers: London, 1998. (11) Metaxas, A. C.; Meredith, R. J. Industrial MicrowaVe Heating; IEE: London, 1983. (12) Sitaram, S. P.; Stoffer, J. O. Microwave Initiated Free Radical Catalyzed Polymerizations: Polystyrene. Part I. Polym. Mater. Sci. Eng. 1993, 69, 382. (13) Stoffer, J. O.; Sitaram, S. P. Microwave Initiated Free Radical Catalyzed Polymerizations: Polystyrene. Part II. Polym. Mater. Sci. Eng. 1994, 71, 55.

ReceiVed for reView September 22, 2009 ReVised manuscript receiVed January 20, 2010 Accepted January 23, 2010 IE901389A