Bitumen Chemical Foaming for Asphalt Paving Applications

Aug 18, 2010 - Bitumen Chemical Foaming for Asphalt Paving Applications. Virginia Carrera, Moises Garcıa-Morales,* Francisco J. Navarro, Pedro Partal...
2 downloads 0 Views 3MB Size
Download PDF
8538

Ind. Eng. Chem. Res. 2010, 49, 8538–8543

Bitumen Chemical Foaming for Asphalt Paving Applications Virginia Carrera, Moises Garcı´a-Morales,* Francisco J. Navarro, Pedro Partal, and Crispulo Gallegos Departamento Ingenierı´a Quı´mica, Facultad de Ciencias Experimentales, Campus “El Carmen”, UniVersidad de HuelVa, 21071 HuelVa, Spain

As an alternative to traditional hot-mix asphalt paving, bitumen foaming technologies have been extensively used in road recycling over the last decades. However, much more work on the use of foam-enhancing chemical agents is needed. This research deals with the influence that temperature and processing protocol exert on the bitumen foaming process promoted by a poly(propylene glycol) functionalized with a polymeric 4,4′diphenylmethane diisocyanate. Likewise, rheological and microstructural characterizations have been carried out on the modified bituminous residues obtained after foaming. From the experimental results obtained, it may be concluded that reactive isocyanate-based polymers can be used to obtain foamed binders with different foamability and stability characteristics. On the other hand, both foaming and bitumen modification are seen to occur simultaneously, as a result of bitumen/polymer reactions involving water. Finally, dynamic shear tests and atomic force microscopy observations reveal very different degrees of modification, which depend on the colloidal features of the as-received bitumen. 1. Introduction Bitumen, a byproduct resulting from crude oil refining, has been largely used as a binder of mineral aggregates in pavement construction. It is composed of asphaltene micelles (solid asphaltene particles covered by a shell of resins) dispersed in a liquid phase, constituted of the remaining resins, along with aromatics and saturates.1,2 Temperature and chemical composition exert a very strong effect on the bitumen microstructure and physical properties. As an alternative to traditional hot-mix asphalt, both low energy consumption and environmental-side-effect-free paving technologies, which involve the mixing of mineral aggregates with a bitumen emulsion or foamed bitumen, have emerged over the last decades. In particular, cold in place recycling, which has become an economical and sustainable paving practice, has been extensively used in the rehabilitation of pavements in many countries throughout the world.3,4 This procedure involves reclaimed asphalt pavement from the existing road, being crashed to the required gradation and mixed with a cold binder. Emulsification and foaming reduce the quantity of bitumen needed, compared to traditional paving, because of an increase in the surface area and, consequently, a reduction in the viscosity. However, a potential advantage of foamed bitumen is a shorter curing time before traffic gets back on the route because foamed bitumen requires less water than an emulsion. The bitumen foaming process, first proposed by Csanyi in the mid 1950s and further improved by a number of oil companies, consists of the injection of water into hot bitumen. As a result, the physical properties of the bitumen are temporarily altered when the injected water, in contact with the hot bitumen, is turned into vapor, which is trapped in thousands of tiny bitumen bubbles.3,5,6 However, relatively little is known about the development of processes involving the use of foam-enhancing agents. On these grounds, the present investigation proposes the employment of a poly(propylene glycol) functionalized with 4,4′diphenylmethane diisocyanate, which provokes a remarkable * To whom correspondence should be addressed. Phone: +34 959 21 82 07. Fax: +34 959 21 93 85. E-mail: [email protected].

increase in the time taken for the foam to collapse and leads to a modified binder showing a significantly improved performance. With the aim to satisfactorily understand the results obtained in this research, the chemistry behind the bitumen foaming process, by using isocyanate-based reactive polymers, should first be detailed. It will also help to achieve a better comprehension of the modification observed further.7,8 In the standard procedure, foaming is promoted by the formation of tiny bubbles of steam after water comes into contact with hot bitumen (above 100 °C and by a sudden pressure reduction), giving rise to foamed bitumen, which resumes its original properties after the foam dissipates. The reactive foaming method herein proposed (in which pressurized air is not required) involves the reaction between water and free -NCO, producing an amine and releasing carbon dioxide, which favors bitumen expansion: R1NCO + H2O f R1NH2 + CO2v

(1)

In the meantime, bitumen modification would result from a twostep mechanism consisting of (a) asphaltene cluster formation, by the reaction of polar groups in the asphaltene micelles and -NCO groups in the reactive polymer: R1NCO + R2OH f R1NHCOOR2

(2)

and (b) further reactions between highly reactive amine groups (due to water addition) and -NCO groups left in the clusters previously formed: R2NCO + R1NH2 f R2NHCONHR1

(3)

The goal of the present work was to gain new insight into the bitumen foaming process promoted by isocyanate-based reactive prepolymers. In addition, a study on the rheology and microstructure of the modified bitumen residues after foaming was carried out. The results obtained may represent a step forward on the development of novel foamed asphalt technology, involving polyurethane-modified bitumen. 2. Experimental Section Four bitumen samples (noted as A-D, 150/200 penetration grade) were used as base materials for foaming. The results of

10.1021/ie101136f  2010 American Chemical Society Published on Web 08/18/2010

Ind. Eng. Chem. Res., Vol. 49, No. 18, 2010 Table 1. Penetration, Ring-and-Ball Softening Temperature, and Gaestel Colloidal Index Values for the Different Neat Bitumen Samples Studied bitumen bitumen bitumen bitumen A B C D penetration [dmm] ring-and-ball softening point [°C] Gaestel colloidal index (Ic)

145 41.5 0.26

160 39.5 0.44

111 45.5 0.46

165 38 0.33

selected technological tests carried out on these samples and the corresponding values of the Gaestel colloidal index9,10 are included in Table 1. The polymer used was a poly(propylene glycol) (PPG) functionalized by polymeric MDI (4,4′-diphenylmethane diisocyanate), henceforth, MDI-PPG. This polymer was synthesized by the reaction of PPG (Alcupol D-0411, donated by Repsol YPF, Spain) and polymeric MDI (supplied by Dow Chemical, Spain), selecting a PPG/MDI molar ratio of 1:3, in a N2 atmosphere, at 40 °C, for 48 h and under agitation. The polymer presents an average Mw of 2800 g/mol, a polydispersity (Mw/ Mn) of 1.33, and an average functionality of 2.8. Blends of every bitumen type and 4 wt % MDI-PPG were prepared. With this aim, four blends were prepared at different processing conditions, which resulted in a combination of two mixing times (15 min and 1 h) and two processing temperatures (90 and 120 °C). Such processing conditions were selected according to the liquid polymer characteristics and time required by the chemical reaction, as reported by Martin-Alfonso et al.11 Mixing was carried out in a cylindrical vessel (80 mm diameter and 140 mm height) by using an IKA RW-20 stirring device (Germany) equipped with a 45°-pitched, four-bladed turbine rotating at 1200 rpm. Additional samples of every bitumen type were prepared, by setting in an oven (for 24 h, at 90 and 120 °C, respectively) blend that had been previously mixed, for 1 h, at the same temperatures. Finally, foaming tests at 90 and 120 °C, depending on the previous processing temperature, were conducted on the resulting modified bitumen samples. With this aim, 2 wt % water was added to the modified bitumen. Samples were stirred, at 500 rpm, for 30 s. Agitation was then stopped, and the foaming process was followed. Readings taken by the

8539

laser device DLS-C 15 distance laser sensor (Dimetix AG, Zurich, Switzerland) determined the evolution of the foam height with time. Once the foaming tests were concluded, bitumen residues were kept for further characterization. For the sake of brevity, samples prepared according to the procedures outlined above will be, henceforth referred, to as “15-min”, “1-h”, and “24-h” samples. For the sake of clarity, legends shown in the different figures include both processing time and, as a superscript, processing and foaming temperatures. Viscous flow measurements, at 60 °C, were carried out in a controlled-stress MARS II rheometer (Thermo Haake, Hannover, Germany), using a plate-and-plate geometry. Frequency sweep tests, at 5 and 60 °C, at stress values inside the linear viscoelastic range of the sample, were conducted in the same rheometer, between 0.03 and 100 rad/s. Average values of, at least, two replicates have been used. Modulated differential scanning calorimetry (MDSC) was performed with a Q-100 (TA Instruments, New Castle, DE). Samples of 5-10 mg were subjected to the same testing procedure: temperature range between -80 and +100 °C; heating rate of 5 °C/min; amplitude of modulation of (0.5 °C, period of 60 s; nitrogen as the purge gas, with a flow rate of 50 mL/min. In order to provide the same recent thermal history, all of the samples were placed in hermetic aluminum pans, for 24 h, at room temperature before measurement. Bitumen SARA fractions (used in the calculation of Gaestel colloidal index values) were determined by thin-layer chromatography coupled with a flame ionization detector (TLC/FID), using an Iatroscan MK-6 analyzer (Iatron Corp. Inc., Tokyo, Japan). Elution was performed in hexane, toluene, and dichloromethane/methanol (95:5), following the procedure outlined elsewhere.12 The microstructural characterization of the samples was carried out by means of atomic force microscopy (AFM), with a atomic force microscope connected to a Nanoscope IV scanning probe microscope controller (Digital Instruments, Veeco Metrology Group Inc., Santa Barbara, CA). All of the images were acquired in tapping mode at 50 °C. The samples

Figure 1. Evolution of foamed bitumen ER with time, at two different testing temperatures (90 and 120 °C), for differently modified bitumen B samples.

8540

Ind. Eng. Chem. Res., Vol. 49, No. 18, 2010

were prepared by heat casting, a method that causes a negligible effect on the material morphology if compared to solvent casting.13 3. Results and Discussion 3.1. Reactive Polymer-Modified Bitumen Foaming. Foaming tests, at 90 and 120 °C, were conducted on the different binders resulting from reactive modification of the four neat bitumen samples used in this study. Because of its high modification capability (see section 3.2), bitumen B was selected as a “model” system to illustrate the process of reactive foaming. Thus, Figure 1 presents the evolution with the time of the expansion ratio (ER) for the “15-min”, “1-h”, and “24-h” bitumen B samples, at both testing temperatures. The ER, related to the viscosity of the foam and, consequently, to the quality of the binder dispersion,4 has been calculated as the ratio of the volume of foam relative to the original volume of reactive polymer-modified bitumen. At 90 °C, large differences among the three samples studied can be noticed. Thus, concerning the “15-min” sample, ER suddenly increases up to a maximum value of 5 and then dramatically decreases. This evolution may be explained by taking into consideration the large -NCO group content left in the reactive polymer, after 15 min of processing, that favors the fast formation of bubbles (the first increase in ER) within a weak bituminous matrix, which then quickly collapses. On the other hand, bitumen processing for 1 h and, mainly, for 24 h results in marked drops in both the foamability and foaming rate. Thus, regarding the maximum foamability, ERmax, the lowest values are obtained for the “24-h” sample, ERmax ) 3.5. As for the foaming rate, the slopes of the curves during the expansion step significantly decrease as the bitumen processing time increases. These results are related to a lower quantity of free -NCO groups available after the reaction between MDIPPG and some of the bitumen compounds [reaction (2)], as the processing time increases, and the consequent reduction in the volume of carbon dioxide released after water addition [reaction (1)]. On the other hand, both “1-h” and “24-h” samples display quite stable foams over the time interval tested. Moreover, even if the experimental time scale is extended, the foamed bitumen obtained does not resume its original volume. As a proof of this assertion, the inset in Figure 1 shows that, for the “24-h” sample, ER still remains above 2 after 300 min of testing time. In addition to the lower quantity of -NCO groups available in the “24-h” sample, its lower foaming rate should further be explained on the basis of the initial viscosity of the modified binder to be foamed. Thus, Figure 2 shows that “24-h” modified bitumen samples present higher viscosities than “1-h” modified bitumen samples, as a consequence of the formation of new microstructures with enhanced strength.11 Moreover, these microstructures would favor the stability of the resulting foam and would reduce the bitumen foaming rate. On the contrary, the three samples studied present the same general pattern when foaming tests are carried out at 120 °C. Thus, ER always shows a rapid increase within the first 1-2 min, up to a maximum ER value, which is followed by an exponential decay. Once again, the highest ERmax values are reached for the “15-min” and “1-h” samples (between 3.75 and 4), whereas a slightly lower value is noticed for the “24-h” sample (about 3.25). Also, ER values, for both the “15-min” and “1-h” samples, decrease down to 1 (initial value) after 20 min, showing the “24-h” sample a value of about 1.5 at that elapsed time.

Figure 2. Viscous flow curves, at 60 °C, for neat bitumen B and “1-h” and “24-h” modified bitumen B samples before and after the foaming process.

When foaming tests, at 90 and 120 °C, for the “15-min” samples are compared, a delay in the ER overshoot with a decrease in the foaming temperature is observed. In other words, a longer elapsed time before foam collapse starts is apparent at 90 °C. These findings agree with the results previously reported for standard bitumen foaming.3 Bubbles collapse when the elongation limit of the covering bitumen film is exceeded. The bitumen viscosity at 90 °C is higher than that at 120 °C. Accordingly, such an elongation limit would be more difficult to exceed in a matrix with a higher viscosity, that is, at lower temperature. Hence, a higher number of unbroken bubbles accumulate continually, giving rise to both higher values of ERmax and the elapsed time at the overshoot.3 On the other hand, the behavior observed may also be the result of water loss, due to evaporation, when it comes into contact with hot bitumen at 120 °C (above the water boiling point). This may lead to a reduction in the volume of carbon dioxide produced. Finally, bitumen foams generally show values of ERmax either very close or above 4, which has been reported to be the minimum value for appropriate mixing in the paving practice.4 Also, an additional benefit of this reactive foaming, in relation to the traditional practice, would be the longer elapsed time until the foam collapses, which would turn out, when required, in more prolonged periods of lay-down and compaction operations. So, using reactive isocyanate-based polymers may become a flexible way to obtain a wide variety of foamed binders with different foamability and stability characteristics, by merely changing the processing time and foaming temperature. With the aim of evaluating the applicability of the different binders obtained, the viscous behavior, at 60 °C, of the bituminous residues obtained after foaming tests, at 90 and 120 °C (in other words, the resulting material after the foam collapses), has been studied. The results shown in Figure 2 reveal that foaming produces a very significant viscosity enhancement [more than 1 order of magnitude, as a consequence of reactions (1) and (3)], as compared to nonfoamed samples [modification resulting from reaction (2)], and more than 2 decades when compared to the neat bitumen. On the other hand, Figure 2 also hints that samples cured for 24 h would lead to binding materials that could show a higher resistance to permanent deformation, at high in-service temperatures, than those foamed just after mixing both bitumen and polymer. Furthermore, a similar modification degree is obtained at both temperatures of processing and foaming. In this sense, the

Ind. Eng. Chem. Res., Vol. 49, No. 18, 2010

8541

Figure 3. Frequency dependence of the linear viscoelasticity functions G′ and G′′, at 60 °C, for neat bitumen samples A and B and “1-h” and “24-h” modified binder residues after foaming at 90 °C.

Figure 4. Frequency dependence of the loss tangent, at 60 °C, for the different neat bitumen samples studied and “1-h” and “24-h” modified binder residues after foaming at 90 °C.

residues obtained after water addition at 120 °C show slightly lower viscosities than those foamed at 90 °C. This fact would confirm that bitumen foaming, at 120 °C, mainly takes place by CO2 released from reaction (1) and, probably, water vapor. 3.2. Rheology and Microstructure of Foamed Bitumen Residues. It has been seen that reactive MDI-PPG-modified bitumen foaming, at 90 and 120 °C, leads to a binding material with highly enhanced viscosity at 60 °C, a fact that may improve its performance at high in-service temperatures. With the aim to gain further insight into the origin of the behavior observed, a more detailed study has been conducted on the corresponding residues and the results compared to those of neat bitumen samples. Frequency sweep test results, at 60 °C, carried out on binder residues, after foaming, at 90 °C, of neat bitumen samples A and B, are shown in Figure 3. Neat bitumen always shows a predominant viscous behavior at 60 °C, with G′′ being proportional to ω and G′ to ω2 at low frequencies, as expected within the terminal viscous region. Nevertheless, the residue behavior strongly depends on the type of bitumen used. Thus, reactive foaming of bitumen A yields much more remarkable increases in G′ than in G′′, although both samples studied still show a predominant viscous response. Also, modified bitumen A samples seem to be quite insensitive to the processing time. On the contrary, outstanding differences arise as a result of bitumen B reactive foaming. Long modified bitumen processing (i.e., 1 h) leads to a remarkable increase in both viscous and elastic moduli, with the latter increasing faster than the former and, consequently, dampening the differences between them (enhanced elastic response). Moreover, the viscous and elastic moduli become closer and parallel in the high-frequency region. Thus, their evolution with frequency is described by two powerlaw functions: G′(ω) ) k1ωn and G′′(ω) ) k2ωn, with the exponent n being approximately 0.5. According to De Rosa and Winter,14 this is related to the onset of a gel-like behavior (criterion for detection of the gel point). Processing for 24 h reduces the differences between both linear viscoelasticity functions further, yielding similar values of G′ and G′′ in a wide frequency range. These results suggest the development of a new microstructure, which enhances material solidlike properties. Rheology testing also evidences different levels of binder modification as a function of the type of bitumen used. In order to emphasize the effect of the bitumen colloidal nature on the resulting viscous/elastic property ratio, Figure 4 presents the

frequency dependence of the loss tangent, measured at 60 °C, corresponding to all bitumen residues that were foamed at 90 °C. Samples of neat bitumen display high values of tan δ (pronounced viscous response at 60 °C), which remarkably decrease after reactive foaming. However, such a decrease is much more important for bitumen C and, mainly, B, where tan δ clearly tends to a constant value as the frequency increases. Moreover, for “24-h” samples of bitumen samples B and C, this “plateau” reaches a value of 1 in a wide frequency range (the same G′ and G′′ values). The Gaestel colloidal index (Ic)9,10 may be considered an important parameter to estimate what type of bitumen is more likely to undergo a higher degree of modification because a higher value of this parameter indicates a more significant bitumen colloidal nature. This index, expressed in terms of SARA fractions, can be written as follows: Ic )

saturates + asphaltenes aromatics + resins

(1)

Colloidal index data shown in Table 1 demonstrate that bitumen samples with the highest values of Ic (B and C) give rise to the most remarkable elasticity enhancement (Figure 4). The results so far reported would imply that when combined with a bitumen sample presenting a high value of Ic, the employment of 4 wt % MDI-PPG as a foaming agent may help relieve hightemperature-related distresses in pavements, such as the wellknown permanent deformation.15,16 However, although the colloidal indexes calculated (see Table 1) suggest some differences among the colloidal structure of the four neat bitumen samples, a method that is only based on the chemical affinity between bitumen fractions and different organic solvents cannot give a detailed idea on the actual bitumen microstucture. Instead, AFM tests carried out on the same bitumen samples,17,18 at 30 °C, clearly showed the socalled “flake-like” dispersion and “bee” structure for bitumen samples A and B, respectively (i.e., neat bitumen B presented a well-developed colloidal microstructure). Taking this fact into consideration, Figure 5 displays AFM micrographs, at 50 °C and in tapping mode, corresponding to neat samples and residue from “24-h” modified bitumen B (i.e., after foaming) samples. The bitumen microstructure, pictured as solid particles of asphaltenes (black and white streaks) covered by a solid shell of resins (light areas) surrounded by the molten maltenic matrix

8542

Ind. Eng. Chem. Res., Vol. 49, No. 18, 2010

Figure 5. AFM micrographs, at 50 °C, for neat bitumen samples A and B and a “24-h” modified bitumen B residue after foaming at 90 °C. Window size 30 × 30 µm.

Figure 6. Frequency dependence of the linear viscoelasticity functions, G′ and G′′, at 5 °C, for neat bitumen samples C and D and “1-h” and “24-h” modified binder residues after foaming at 90 °C.

(darkest areas),13 matches the well-known bitumen colloidal model.1,2 However, large differences between both systems can be appreciated. As compared to neat bitumen, the “24-h” residue presents much larger asphaltenic regions, as a consequence of the reactions already detailed, which almost vanish for neat bitumen at this testing temperature. The high level of interactions among asphaltenic micelles, at 50 °C, in the residue from modified bitumen B would explain the remarkable modification degree attained after reactive foaming (Figures 3 and 4). The previously observed increase in the viscoelastic moduli (and viscosity) might be attributed to the development of a more compact microstructure originated by the formation of covalent bonds between the polymer and certain bitumen groups (present in asphaltenes, resins, and even aromatics), as a consequence of the reaction described above [reactions (1)-(3)].18 Residue rheological properties were also evaluated at low temperature. Thus, frequency sweep tests, at 5 °C, were conducted on bitumen residues after foaming at 90 °C. The results obtained for bitumen samples C and D are presented in Figure 6. As can be observed, neat bitumen still presents a predominant viscous response at 5 °C, with G′′ being higher than G′ in nearly the entire frequency range studied, although a crossover between both linear viscoelastic functions is noticed at very high frequencies. Once again, the binder viscoelasticity is strongly dependent on the bitumen colloidal nature. Hence, bitumen D (with the lowest colloidal index value; Table 1) displays a notable increase in both moduli (for “1-h” and, mainly, “24-h” samples), but with G′′ remaining higher than G′. Regarding bitumen C (with the highest colloidal index value; Table 1), G′ increases much faster than G′′, which shifts the crossover frequency to the lowest values found in this study, yielding values of the elastic modulus higher than those of the

Figure 7. Frequency dependence of the loss tangent, at 5 °C, for the different neat bitumen samples studied and “1-h” and “24-h” modified binder residues after foaming at 90 °C.

viscous one in the entire frequency range tested. Also, reactive foaming causes a pronounced decrease in the slope of both G′ and G′′ versus frequency curves, more significant for “24-h” samples. Moreover, the evolution of the loss tangent with frequency, presented in Figure 7, also highlights the effect of the bitumen colloidal nature on the viscous/elastic property balance for the four types of bitumen used. Thus, those bitumen samples showing the highest colloidal indexes present the lowest values of tan δ. Reactive foaming increases bitumen elastic features, above all for bitumen samples B and C (those with the highest values of Ic; Table 1) after 24 h of processing, for which tan δ remains below 1 in the whole frequency window tested. The decrease in the slope above, resulting from bitumen/ polymer reactions, might suggest a delay in the appearance of the glassy region, in which the binder presents a very high brittleness.19,16 MDSC gives additional information concerning the waterinvolved reactions between polymer and bitumen previously described. Reversing the heat-flow thermograms (see Figure 8) reveals the existence of a pronounced glass transition in polymeric MDI, located at about -50 °C (curve “a”), which moves up to -25 °C (curve “b”) after its reaction with PPG. On the other hand, reversing the thermal curve for neat bitumen B (curve “c”) reveals a large glass transition event, centered at -30 °C.20 Such a thermal event would arise from the overlapping of two glass transitions at about -39 and -14 °C, as confirmed by the inset in Figure 8, which displays the evolution of the heat capacity derivative at low temperature for this neat bitumen. Reactive foaming at 90 °C of bitumen B (curve “d”), processed during 24 h, shifts both thermal events to lower temperatures, as shown by the inset in Figure 8. Hence, MDSC

Ind. Eng. Chem. Res., Vol. 49, No. 18, 2010

8543

Literature Cited

Figure 8. Reverse heat-flow curves for MDI, MDI-PPG, bitumen B, and residue of “24-h” bitumen B. Inset: Heat capacity temperature derivative for neat and residue of “24-h” bitumen B.

results seem to support the above improvement in material properties at low in-service temperatures suggested by rheological tests at 5 °C (Figures 6 and 7). Finally, the endotherm at 0 °C, due to the melting peak of nonreacted water, can be noticed. This fact should be considered for an eventual product optimization. 4. Conclusions The use of PPG functionalized with a polymeric MDI as a foam-enhancing agent for bitumen has been evaluated. Foaming tests, conducted on samples of four different bitumen types, reveal that reactive isocyanate-based polymers may be satisfactorily used to obtain a variety of foamed bituminous binders with different foamability and foam stability before collapse, by merely changing processing time and foaming temperature. On the other hand, reactive foaming with MDI-PPG seems to be a complex process, promoted by the formation of carbon dioxide as a result of reactions between free -NCO groups and water. Simultaneously, reactions involving asphaltene micelles lead to a binder with improved viscoelastic properties, at both high and low in-service temperatures. Finally, the results obtained demonstrate that very different modification levels can be achieved, depending on the bitumen colloidal nature. Precisely, the Gaestel colloidal index values and AFM observations revealed that a larger degree of modification is obtained with neat bitumen samples having a welldeveloped asphaltene-rich colloidal microstructure. Acknowledgment This work is part of a research project sponsored by a MECFEDER programme (Research Project MAT2007-61460). The authors gratefully acknowledge its financial support.

(1) Lesueur, D.; Gerard, J. F.; Claudy, P.; Letoffe, J. M.; Planche, J. P.; Martin, D. A structure-related model to describe asphalt linear viscoelasticity. J. Rheol. 1996, 40, 813. (2) Lesueur, D. The colloidal structure of bitumen: Consequences on the rheology and on the mechanisms of bitumen modification. AdV. Colloid Interface Sci. 2009, 145, 42. (3) He, G.; Wong, W. Decay properties of the foamed bitumens. Constr. Build. Mater. 2006, 20, 866. (4) Jenkins, K. J.; Molenaar, A. A. A.; de Groot, J. L. A.; van de Ven, M. F. C. Developments in the uses of foamed bitumen in road pavements. HERON 2000, 45, 167. (5) Muthen, K. M. Foamed Asphalts Mixes. Mix Design Procedure. Contract Report CR-98/077. Prepared by CSIR TRANSPORTEK, Pretoria, South Africa, 1998. (6) Jenkins, K.; van de Ven, M. Guidelines for the mix design and performance prediction of foamed bitumen mixes. 20th South African Transport Conference “Meeting the Transport Challenges in Southern Africa”. 2001. (7) Martı´n-Alfonso, M. J.; Partal, P.; Navarro, F. J.; Garcı´a-Morales, M.; Gallegos, C. Role of water in the Development of New IsocyanateBased Bituminous Products. Ind. Eng. Chem. Res. 2008, 47, 6933. (8) Segura, D. M.; Nurse, A. D.; McCourt, A.; Phelps, R.; Segura, A. Chemistry of polyurethane adhesives and sealants. In Handbook of adhesiVes and sealants; Cognard, P., Ed.;Elsevier Ltd.: New York, 2005; Chapter 3. (9) Gaestel, C.; Smadja, R.; Lamminan, K. A. Contribution a la connaissance des proprietes des bitumes routiers. ReV. Gen. Routes Aerodromes 1971, 466, 85. (10) Baginska, K.; Gawel, I. Effect of origin and technology on the chemical composition and colloidal stability of bitumens. Fuel Process. Technol. 2004, 85, 1453. (11) Martı´n-Alfonso, M. J.; Partal, P.; Navarro, F. J.; Garcı´a-Morales, M.; Bordado, J. C. M.; Diogo, A. C. Effect of processing temperature on the bitumen/MDI-PEG reactivity. Fuel Process. Technol. 2009, 90, 525. (12) Eckert, A. The application of Iatroscan-technique for analysis of bitumen. Pet. Coal 2001, 43, 51. (13) Masson, J. F.; Leblond, V.; Margeson, J. Bitumen morphologies by phase-detection atomic force microscopy. J. Microsc. 2006, 221, 17. (14) De Rosa, M. E.; Winter, H. H. The effect of entanglements on the rheological behaviour of polybutadiene critical gels. Rheol. Acta 1994, 33, 220. (15) Garcı´a-Morales, M.; Partal, P.; Navarro, F. J.; Martı´nez-Boza, F. J.; Gallegos, C.; Gonza´lez, N.; Gonza´lez, O.; Mun˜oz, M. E. Viscous properties and microstructure of recycled EVA modified bitumen. Fuel 2004, 83, 31– 38. (16) Navarro, F. J.; Partal, P.; Martı´nez-Boza, F. J.; Valencia, C.; Gallegos, C. Rheological characteristics of ground tire rubber-modified bitumens. Chem. Eng. J. 2002, 4005, 1–9. (17) Carrera, V.; Partal, P.; Garcı´a-Morales, M.; Gallegos, C.; Pa´ez, A. Influence of bitumen Colloidal Nature on the Design of Isocianate-Based Bituminous Products with Enhanced Rheological Properties. Ind. Eng. Chem. Res. 2009, 48, 8464. (18) Carrera, V.; Garcı´a-Morales, M.; Partal, P.; Gallegos, C. Novel bitumen/isocyanate-based polymer formulations for the paving industry. Rheol. Acta 2010, 49, 563-572. (19) Lu, X.; Isacsson, U.; Ekblad, J. Low-temperature properties of styrene-butadiene-styrene polymer modified bitumens. Constr. Build. Mater. 1998, 12, 405–414. (20) Masson, J. F.; Polomark, G. M.; Collins, P. Time-dependent microstructure of bitumen and its fractions by modulated differential scanning calorimetry. Energy Fuels 2002, 16, 470.

ReceiVed for reView May 22, 2010 ReVised manuscript receiVed July 16, 2010 Accepted July 20, 2010 IE101136F