Investigation of the Rheological Properties and Ageing Susceptibility of Bitumen Bio-Modified with Spent Coffee Grounds

Abstract

The present study concerns the utilization of spent coffee grounds (SCGs) as an alternative bio-based modifier for a petroleum-based penetration grade 70/100 bitumen at 5%, 10% and 15% by weight of bitumen. The conventional properties of the binders were examined with a series of penetration, ring and ball, elastic recovery, dynamic viscosity and storage stability tests. Their rheological properties were assessed with a Dynamic Shear Rheometer. The aforementioned tests were conducted before and after applying a short-term ageing protocol to quantify the ageing susceptibility of the binders using different rheological ageing metrics. Furthermore, a statistical analysis was conducted to discover whether any correlations exist between the conventional and rheological properties of the binders. It was observed that spent coffee grounds can be incorporated into bitumen at an optimal content of up to 5% without downgrading the binder’s rheological properties or its structural integrity. Additionally, the bio-modifier slightly improved the ageing resistance of bitumen. Finally, the ring and ball test’s results had the strongest correlation with the DSR findings.

1. Introduction

In recent years, the rising economic and environmental costs of petroleum-based products have led researchers to investigate bio-based products as potential alternative construction materials for the conventional petroleum-derived bitumen in the pavement construction industry [1,2]. When these bio-materials replace part of the conventional bitumen, they act as bio-modifiers. Such materials may be derived from different sources, such as plant-based or animal-based organisms, may have various structures (powders, fibres, oils, etc.) and may require different processing methods to be successfully utilized. Such materials can be swine manure, rice husk, algae, waste wood, coconut waste, soybean, etc. [3,4,5,6,7,8]. In most cases, they are at the end of their product lifespan (EOL), such as in the case of waste cooking oil, biodiesel residue and spent coffee grounds [9,10,11,12,13,14], which after their useful life may end up in landfills, causing leaching pollution, or are burnt, intensifying the already-established problem of the greenhouse effect. Therefore, their incorporation into the pavement industry would not only extend their product life but also potentially reduce the hazards related to their disposal on a global scale. One of the main advantages of bio-modifiers over conventional bitumen modifiers is that they are abundant, since the sources where they can be drawn from are numerous. Each bio-modifier aims to alter specific properties of bitumen. For instance, soybean-derived rejuvenators have been found to improve the fatigue and low-temperature performance of the asphalt mixture [15,16,17]. Certain bio-materials, such as lignin and corn stalk-derived hydrochar, have been utilized to improve bitumen and asphalt mixture performance in the high-temperature domain [18,19]. Certain bio-materials, such as date seed ash, have also been investigated via the surface free energy method as potential candidates to improve bitumen–aggregate adhesion [20]. This improvement could moderate stripping and thus limit moisture damage [21]. Many bio-derived materials, such as pyrolysis-derived biochar, can be incorporated into bitumen to reduce its ageing rate by acting as free radical scavengers [22]. However, it should be noted that not all bio-materials and concentrations are beneficial from a performance point of view for all the properties of bitumen or asphalt mixtures. For instance, it has been reported that algae-derived bio-oil can reduce the bitumen–aggregate interfacial forces, which will cause moisture damage, as explained above, in addition to a possible conflict with the biofuel production industry due to its significant calorific value [23]. Furthermore, although soybean oil-containing microcapsules promote the self-healing ability of asphalt mixtures without compromising their thermal stability, they may cause a deterioration of its resistance against moisture and permanent deformation [24].

Taking the above into consideration, this study aims to investigate the potential of spent coffee grounds, an end-of-life material with no conflict with the food or biofuel industry, as a bio-based modifier for bitumen through a series of testing protocols to identify their conventional and rheological properties. This investigation is part of a series of studies that aim to offer a framework for the proper use of bio-materials in the pavement construction industry for both academic and industrial stakeholders alike.

2. Materials and Methods

2.1. Materials

2.1.1. Bitumen

A penetration grade 70/100 bitumen was used as the reference binder and the basis for producing the spent coffee ground-modified binders. The conventional properties of the 70/100 bitumen were determined in previous studies [5] and are given in

Table 1. Conventional properties of 70/100 reference bitumen [5].

Property	Specification	Value	Reference
Penetration (dmm)	EN 1426	79	[25]
Softening Point (°C)	EN 1427	49.8	[26]
Elastic Recovery (%)	EN 13398	11	[27]
Force Ductility (J/cm2)	EN 13703	0.844	[28]
Dynamic Viscosity (Pa.s)	EN 13302	Various 1	[29]

¹ Dynamic viscosity test was carried out at various temperatures. Results are given in Section 3.1.

.

2.1.2. Spent Coffee Grounds (SCGs)

The spent coffee grounds were dried in the oven overnight at 110 °C to remove their moisture before mixing them with the bitumen as a solid-state bio-modifier using a high-shear mixer of 2500 rpm at 160 °C for 30 min. SCGs were incorporated into bitumen as a bio-modifier for physical modification due to their ability to modify (i.e., alter) certain properties of bitumen. A total of three bio-modified binders were produced with 5%, 10% and 15% SCGs by weight of bitumen, denoted as SCG-5, SCG-10 and SCG-15, respectively. The bio-modifier was observed under a Confocal Microscope Leica DCM 8 (Figure 1), with the mean particle size varying between 250 and 500 μm. The particles can be characterized as irregularly and angularly shaped, while their colour is dark brown.

2.2. Methods

2.2.1. Conventional Properties

The conventional properties of the reference bitumen and the bio-modified binders were examined via the penetration (EN 1426) [25], softening point (EN 1427) [26], elastic recovery (EN 13398) [27], dynamic viscosity (EN 13302) [29] and storage stability tests (EN 13399) [30]. In particular, the dynamic viscosity test was carried out with a Brookfield viscometer using a No. 27 spindle at 20 rpm for temperatures ranging between 130 °C and 180 °C with steps of 10 °C, as a way of covering the usual mixing and compaction temperatures. Based on the dynamic viscosity results, the mixing temperature was calculated in accordance with previous works [5]. More specifically, an exponential, non-linear regression was carried out on the dynamic viscosity results to determine the exact temperature at which the dynamic viscosity of the unaged binders is equal to 0.2 Pa.s (T_mixing).

2.2.2. Rheological Properties

Multiple frequency sweeps were carried out for temperatures between 10 °C and 70 °C with steps of 10 °C in order to investigate the rheological properties of the binders. More specifically, the frequencies ranged from 0.01 Hz to 10 Hz. The temperatures from 10 °C to 40 °C were tested with 8 mm DSR parallel plates, whereas the 50–70 °C temperature domain was assessed with 25 mm DSR parallel plates. For each type of binder, two replicates were examined for every geometry. Additionally, the rutting and fatigue cracking factors were determined at 1.59 Hz (10 rad/s) to examine the rutting and fatigue cracking resistance of bitumen at 50 °C and 20 °C, respectively. This specific loading frequency corresponds to a traffic speed of 90 mph. Last but not least, the high critical temperature was calculated for each binder based on the rutting factor results to assess the resistance of bitumen against permanent deformation phenomena [31].

2.2.3. Ageing Protocol

The Rolling Thin-Film Oven Test (RTFOT) was conducted in accordance with EN 12607-1 as a way of simulating the short-term ageing of bitumen, which corresponds to the ageing that bitumen undergoes during the mixing and compaction phases [32].

2.2.4. Ageing Susceptibility

Two rheological ageing indexes were considered to determine the impact that spent coffee grounds have on the ageing susceptibility of bitumen. More specifically, they were calculated based on the rutting (RF) and fatigue cracking (FCF) factors (Equations (1) and (2)).

$${\mathrm{A}\mathrm{I}}_{\mathrm{R}\mathrm{F}}={\displaystyle \frac{{\mathrm{G}\ast /\mathrm{s}\mathrm{i}\mathrm{n}\mathsf{\delta }}_{\mathrm{R}\mathrm{T}\mathrm{F}\mathrm{O}\mathrm{T}}}{{\mathrm{G}\ast /\mathrm{s}\mathrm{i}\mathrm{n}\mathsf{\delta }}_{\mathrm{U}\mathrm{n}\mathrm{a}\mathrm{g}\mathrm{e}\mathrm{d}}}}$$

(1)

$${\mathrm{A}\mathrm{I}}_{\mathrm{F}\mathrm{C}\mathrm{F}}={\displaystyle \frac{{\mathrm{G}\ast \times \mathrm{s}\mathrm{i}\mathrm{n}\mathsf{\delta }}_{\mathrm{R}\mathrm{T}\mathrm{F}\mathrm{O}\mathrm{T}}}{{\mathrm{G}\ast \times \mathrm{s}\mathrm{i}\mathrm{n}\mathsf{\delta }}_{\mathrm{U}\mathrm{n}\mathrm{a}\mathrm{g}\mathrm{e}\mathrm{d}}}}$$

(2)

2.2.5. Correlation Statistical Analysis

In order to observe whether any correlations exist between the conventional and rheological properties of the reference and the bio-modified binders, the R² values were calculated for the respective test results for the unaged and the RTFOT-aged samples. Therefore, a total of 8 observations were considered to construct a 6 × 6 matrix. A high R² value suggests a strong correlation between any given couple of x and y data sets.

3. Results and Discussion

3.1. Conventional Properties

Table 2. Conventional properties of the unaged SCG-modified binders.

	Penetration (dmm)	Softening Point (°C)	Elastic Recovery (%)	ΔTS.P. (°C)
70/100	79	49.8	11	-
SCG-5	53	50.2	5	4.0
SCG-10	51	50.6	5	6.8
SCG-15	47	51.0	5	Not Applicable 1

¹ The test was unsuccessful as the steel balls pierced through the samples.

and

Table 3. Conventional properties of the RTFOT-aged SCG-modified binders.

	Penetration (dmm)	Softening Point (°C)	Elastic Recovery (%)
70/100	60	52.2	8
SCG-5	32	52.8	5
SCG-10	30	53.4	5
SCG-15	30	56.6	Not Applicable 1

¹ The test was unsuccessful as the samples failed to reach 200 mm of elongation.

show the results of the conventional properties of the reference bitumen and the bio-modified binders before and after short-term ageing. A gradual decline in penetration values, which signifies the stiffening effect of SCGs on bitumen, can be observed. It should be noted that the reference bitumen showed a change in penetration grade category from 70/100 to 50/70 in the case of SCG-5 and SCG-10 and to 35/50 in the case of SCG-15 (47 dmm). The softening points demonstrated the opposite trend, which is in alignment with the penetration test findings. Additionally, the elastic recovery of bitumen was not improved after the bio-modification, since all of the samples had 5% elastic recovery. However, it is noteworthy that the SCG-15 binder failed to reach the full elongation of 200 mm after being subjected to the short-term ageing protocol. This means that excessive bio-modifier contents may downgrade the cohesion of bitumen by making it too brittle. Furthermore, the SCG-5 binder remained relatively stable when stored at elevated temperatures for extended periods of time since the softening point difference between the two parts of the aluminum tube was equal to 4.0 °C. SCG-10 showed segregation phenomena with a ΔT_S.P. of 6.8 °C, while SCG-15 failed to give results due to the fact that the steel balls managed to pierce through the samples of the bottom part of the aluminum tube, which is linked to the aforementioned cohesion loss [22]. The short-term ageing protocol caused the bitumen to become stiffer, as proven by the reduced penetration values and the increased softening points across all binders.

Table 4. Brookfield viscometer results (unaged).

	Dynamic Viscosity (Pa.s)						Tmxing (°C)
	130 °C	140 °C	150 °C	160 °C	170 °C	180 °C
70/100	0.350	0.250	0.175	0.125	0.100	0.075	146
SCG-5	0.475	0.350	0.250	0.200	0.125	0.100	156
SCG-10	0.525	0.350	0.250	0.200	0.125	0.100	159
SCG-15	0.550	0.350	0.250	0.200	0.150	0.100	161

and

Table 5. Brookfield viscometer results (RTFOT).

	Dynamic Viscosity (Pa.s)
	130 °C	140 °C	150 °C	160 °C	170 °C	180 °C
70/100	0.475	0.325	0.200	0.150	0.100	0.075
SCG-5	0.675	0.450	0.300	0.200	0.150	0.100
SCG-10	0.675	0.450	0.300	0.200	0.150	0.125
SCG-15	0.775	0.500	0.325	0.225	0.150	0.125

present the dynamic viscosity and mixing temperature results from the Brookfield viscometer. The main finding is that an increasing SCG content led to increasing dynamic viscosity values, especially after short-term ageing. The increment was equally pronounced for the unaged and RTFOT-aged samples and was more prominent in the 130–160 °C temperature spectrum. The mixing temperature also became higher, up to 161 °C for the SCG-15 binder compared to 146 °C for the reference bitumen. This verifies that the temperature of 160 °C selected for the mixing protocol was correct.

3.2. Rheological Properties

3.2.1. Master Curves

The stiffening effect of SCG on bitumen was also observed during the rheological investigation of the bio-modified binders (Figure 2 and Figure 3). In particular, a slight increase in the complex shear modulus and a trivial decrease in the phase angle were noticed in the whole frequency spectrum. However, the SCG-15 binder before and after ageing and the SCG-10 binder after ageing manifested a different behaviour in the low-frequency domain. The profile of the phase angle master curves of SCG-10 and SCG-15 could potentially be characterized as parabolic-like. The above frequency spectrum represents the high-temperature domain (namely the 25 mm diameter samples between 50 °C and 70 °C). At such high temperatures, the bitumen tends to soften since they are much higher than its softening point. Therefore, the DSR device may mostly record the friction and the interlocking effect between the spent coffee ground particles instead of the stiffness of the composite material, especially if the bio-modification content is too high or the particle size is excessively large. The interlocking phenomenon is further intensified by the fact that SCG particles are angular in shape (Figure 1). The fact that the minimum content where this anomaly is displayed is lower after ageing (10% instead of 15% by weight of bitumen) is probably related to the fact that the oxidative ageing of bitumen further downgrades its viscoelastic properties. This observation is also linked to the results of the storage stability test, which could not be conducted for SCG-15 due to the serious segregation that caused the steel balls to pierce through the samples, as well as the findings of the elastic recovery test. These conclusions suggest that certain bio-modifiers have a threshold content which, when exceeded, deteriorates the viscoelastic behaviour of bitumen. This threshold seems to be close to 5% in the case of spent coffee grounds.

3.2.2. Black Space Diagrams

The black space diagrams of the reference bitumen and the SCG-5 binder maintained the same form before and after applying the RTFOT ageing protocol (Figure 4 and Figure 5). However, the rest of the binders demonstrated a half-elliptical-like profile, which is in alignment with the master curve findings. When bitumen softens, particle-to-particle friction may cause interlocking phenomena within the matrix, which makes the bitumen seem like it reaches a plateau in G* values. Therefore, black space diagrams showed that an SCG content of 10% or higher should be avoided in order to maintain the homogeneity of the binder without causing cohesion loss.

3.2.3. Rheological Indexes

According to Figure 6, the rutting factor of bitumen slightly improved after the SCG bio-modification. The SCG-5 binder presented almost the same factor as the 70/100 reference bitumen. The difference in rutting factors was more apparent in the case of SCG-10, where the G*/sinδ ratio improved from 12.90 kPa to 15.32 kPa for the unaged samples and from 25.70 kPa to 28.46 kPa for their aged counterparts. The unaged SCG-15 binder followed a gradual increasing trend, but its short-term aged counterpart presented a steep increment to 59.32 kPa. As explained above, this behaviour may not be attributed exclusively to bitumen, since the testing temperature for the rutting factor was 50 °C, which is very close to the softening point of the bitumen, as discussed in Section 3.1. This means that major performance loss is displayed for testing temperatures of 50 °C and above.

The fatigue cracking factors of the reference bitumen and the bio-modified binders are displayed in Figure 7. The resistance of bitumen against fatigue cracking remained almost unchanged for an SCG content up to 10%, especially for the short-term-aged samples. Characteristically, the G* × sinδ ratio increased from 5.02 × 10³ kPa to only 5.32 × 10³ kPa for the RTFOT-aged samples. The SCG-15 binder showed the worst behaviour in terms of fatigue cracking, as proven by 37.1% and 17.7% increments from the reference bitumen for the unaged and aged states, respectively. Nevertheless, since the testing temperature for the fatigue cracking factors was 20 °C, the cohesion of the bio-modified binders with a high SCG content was partially maintained. This was expected, as segregation phenomena between the bitumen and the modifier are more prone to appear at elevated temperatures. In any case, it can be observed that the RTFOT-aged samples displayed worse cracking behaviour due to the stiffening effect when bitumen undergoes oxidative ageing [33].

3.3. Ageing Susceptibility

Both rheological ageing indexes showed that SCGs contributed to a slight reduction in the ageing susceptibility of bitumen for a rheological standpoint, especially in the case of SCG-10 (Figure 8). More specifically, the ageing indexes of the reference bitumen (1.992 for G*/sinδ and 1.422 for G* × sinδ) were reduced down to 1.858 and 1.313 for SCG-10. The smallest ageing susceptibility in terms of fatigue cracking ageing factors was displayed by SCG-15, with a value of 1.221. However, its rutting factor ageing index was also the highest. But neither of these ageing metrics should be considered a proper means of quantifying the ageing susceptibility for this specific binder due to the loss of rheological integrity, as proven by the findings of Section 3.1 and Section 3.2.

3.4. Correlation Statistical Analysis

Table 6. R² values between various bitumen test results.

	Penetration	S.P.	n130 °C	G*/sinδ	G* × sinδ	Tmixing
Penetration	1.00
S.P.	0.55	1.00
n130°C	0.92	0.78	1.00
G*/sinδ	0.39	0.95	0.65	1.00
G* × sinδ	0.67	0.83	0.79	0.70	1.00
Tmixing	0.98	0.87	1.00	0.55	0.68	1.00

displays the R² values between the results of different tests. Among the four conventional properties of the matrix (penetration, softening point, dynamic viscosity at 130 °C, mixing temperature), the softening point correlates the best with the rheological metrics (0.95 and 0.83 with G*/sinδ and G* × sinδ, respectively), followed by dynamic viscosity at 130 °C (0.65 and 0.79, respectively) and, lastly, penetration, which had the weakest correlation. In most cases, the least-related rheological index to the conventional properties was the rutting factor, which is probably attributed to the results of the SCG-15 binder, which were much higher due to the aforementioned interlocking effect, which is more pronounced in the high-temperature domain, namely where the rutting factor is determined, and which therefore acted as an outlier for the established correlations and caused a decline of the R² value.

4. Conclusions

The main conclusions from the present study are as follows:

• Spent coffee grounds had a stiffening effect on bitumen, as displayed by the findings on their conventional and rheological properties.
• A spent coffee ground dosage of 5% can be incorporated into bitumen without negatively affecting its rheological integrity and its fatigue resistance.
• When a bio-modifier content of 5% is exceeded, the rheological behaviour is substantially downgraded, as deduced by the structural changes in the binders according to the master curves and the black space diagrams.
• The ageing susceptibility metrics indicated a slight improvement in the ageing behaviour after the bio-modification up to a bio-modifier content of 10%.
• Ring and ball test results are the ones that correlate the best with the rheological indexes out of all the conventional properties.