Proximate and elemental analysis of lavandin and hydrochars
Table 2 presents proximate and elemental analysis of lavandin and its hydrochars. Moisture content of lavandin sample was initially low (< 3% wt) compared to other biomass [44]. After HTC treatment, moisture content decreased below 1%. Volatile matter in lavandin sample was near 81% and decreased in hydrochars as retention time increased and/or as temperature increased. On the contrary, fixed carbon, initially low for lavandin (8.55%) increased in hydrochars as the retention time and/or the temperature increased. Compensation of volatile matter decrease by fixed carbon increase is represented by the fuel ratio. An increase in fuel ratio indicates that fixed carbon increases and volatile matter decreases. Hence fuel ratio can be used as one of the determining parameters for evaluation of biofuel quality. As an example, the fuel ratio of hydrochars obtained from rice husk and coconut husk were in the range 0.66–0.86 and 0.43–0.62 [45], the fuel ratio of hydrochars obtained from filamentous algae (H. reticulatum) and microalgae (C. vulgaris) were 0.13–0.2 and 0.12–0.16, respectively [46]. Fuel ratio was 0.11 for lavandin and it ranged from 0.3 to 0.69 for hydrochars. Fuel ratio of lavandin hydrochars were on the same order of magnitude than fuel ratio for rice and coconut husks hydrochars [43] meaning that hydrochars were better solid fuels than raw biomass. Evolution of fuel ratio depending on severity factor in presented on Fig. 1 (a), and is compared to fuel ratio from Liu et al. [47] for coconut fibers and eucalyptus leaves hydrochars obtained for the same S:L conditions than the one used in this study and severity factors were ranging from 4.42 to 9.57. Fuel ratios were in the same order of magnitude for lavandin hydrochars, coconut fibers hydrochars and eucalyptus leaves hydrochars. Fuel ratio linearly increased with severity factor for the three hydrochars. The growth rate was the same for lavandin and coconut fibers hydrochars, and it was slightly lower for eucalyptus leaves. Evolution of fuel ratio indicated that at the highest severity factor values, fixed carbon content was the highest and volatile matter content was the lowest. For lavandin hydrochars, sample LH20-260 presented the best fuel properties, compared to other lavandin hydrochars. Ash content was around 7.6% in lavandin and it was low for all lavandin hydrochars (around 5%), indicating a low amount of combustion residues. Regarding ash content, lavandin hydrochars are better fuels than lavandin. From ultimate analysis, carbon content in lavandin hydrochars increased as severity factor increased, while oxygen content decreased as severity factor increased. Nitrogen and hydrogen variations were less noticeable (Fig. 1 (b)). Calculated HHV values ranged from 21.43 MJ kg− 1 for lavandin to 30.77 MJ kg− 1 for lavandin hydrochar. These HHV values were higher than HHV obtained from sewage sludge hydrochar [48, 49] (around 10–15 MJ kg− 1) but were in the same order of magnitude than hydrochars from miscanthus, eucalyptus bark, empty fruit brunches, pine wood meal, lignocellulosic biomass, eucalyptus leaves, coconut fibers [22, 50]. Figure 1 (b) presents evolution of HHV versus severity factor for lavandin hydrochars and coconut fibers and eucalyptus leaves hydrochars from Liu et al. [47] for comparison. HHV values obtained at different severity factors for different kind of biomass and the same S:L initial condition, are of the same order of magnitude and ranged from 25 to 30 MJ kg− 1. Hence, calorific values for lavandin hydrochars are comparable with the ones for coconut fibers and eucalyptus leaves, indicating that lavandin hydrochars could be interesting solid fuel. From SF 4 to SF 5, HHV value remained constant irrespectively of SF value. Between SF 5 and SF 6, HHV value increased from 25 to 30 MJ kg− 1, then above SF 6, HHV value reached a plateau and remained constant whatever the SF value. In the case of lavandin hydrochars, HHV value evolution is directly correlated to carbon content. The described trend was the same for coconut and eucalyptus hydrochars, with an increase of HHV for severity factor from 5 to 6, then above 6 HHV value did not change significantly and remained stable. In conclusion, regarding HHV value, hydrochars obtained for SF 6 seemed to be a good compromise. Figure 1 (c) presents evolution of hydrochar yield as a function of severity factor. Two distinct segments arose: the first one concerned severity factors from 4 to 6 where hydrochar yield strongly decreased with a slope near to 8%/SF unit; the second segment concerned severity factors above 6, where hydrochar yield decrease was slighter with a slope near to 2%/SF unit. Regarding optimal hydrochar yield, a severity factor around 6 seemed to be a good compromise for solid fuel application. Evolution of energy yield as a function of severity factor is presented Fig. 1 (d). In the range of severity factors studied, energy yield of lavandin hydrochars decreased from 68–53%, and the maximum value for energy yield corresponded to the sample obtained at the lowest severity factor value. Compared with energy yields for coconut fibers and eucalyptus leaves hydrochars [41], lavandin, coconut fibers, and eucalyptus hydrochars had energy yields in the same order of magnitude and depicted a linear decrease trend as severity factor increased and the rates of decay were similar. The rate of decay was similar for lavandin and coconut fibers hydrochars, whereas it was higher for eucalyptus leaves hydrochars, indicating that the latter were less interesting biofuels compared to lavandin and coconut fibers.
Table 2
Proximate and ultimate analysis of lavandin sample and its hydrochars. (*) Fuel ratio corresponds to the fixed carbon/volatile matter ratio; (**) HHV was calculated according to Channiwala et al. (2002) [38]; ( ****) Hydrochar yield (%) was calculated as \(\frac{dried hydrochar weight}{dried biomass weight}\times 100\); (***) Energy yield corresponds to\(\frac{{HHV}_{hydrochar}}{{HHV}_{biomass}}\times hydrochar yield\)
|
L
|
LH1-220
|
LH2-220
|
LH4-220
|
LH8-220
|
LH20-220
|
LH4-180
|
LH4-200
|
LH4-240
|
LH20-200
|
LH20-240
|
LH20-260
|
Moisture (%)
|
2.93
|
0.68
|
0.65
|
0.51
|
0.42
|
0.4
|
1.21
|
1.06
|
0.39
|
0.65
|
0.39
|
0.33
|
Volatile matter (%)
|
80.89
|
72.66
|
70.28
|
67.08
|
63.6
|
60.67
|
70.19
|
74.96
|
61.65
|
65.54
|
58.43
|
54.91
|
Fixed carbon (%)
|
8.55
|
21.86
|
24.07
|
27.77
|
31.2
|
34.34
|
23.25
|
19.83
|
33.22
|
28.41
|
35.76
|
37.8
|
Ash (%)
|
7.63
|
4.8
|
5.0
|
4.64
|
4.78
|
4.59
|
5.35
|
4.15
|
4.74
|
5.4
|
5.42
|
6.96
|
Fuel ratio *
|
0.11
|
0.30
|
0.34
|
0.41
|
0.49
|
0.57
|
0.33
|
0.26
|
0.54
|
0.43
|
0.61
|
0.69
|
C (%)
|
50.77
|
60.42
|
61.54
|
64.98
|
67.24
|
70.1
|
58.79
|
61.35
|
69.65
|
65.31
|
71.94
|
71.59
|
H (%)
|
6.9
|
6.81
|
6.31
|
6.27
|
6.28
|
6.51
|
6.48
|
6.21
|
6.75
|
6.82
|
6.65
|
6.04
|
N (%)
|
1.28
|
1.09
|
1.18
|
1.31
|
1.44
|
1.62
|
1.29
|
1.24
|
1.53
|
1.46
|
1.69
|
1.77
|
O (%)
|
41.05
|
31.68
|
30.97
|
27.44
|
25.04
|
21.77
|
33.44
|
31.2
|
22.07
|
26.41
|
19.72
|
20.6
|
HHV (MJ/kg) **
|
21.43
|
25.72
|
25.59
|
27.12
|
28.16
|
29.77
|
24.57
|
25.40
|
29.86
|
27.97
|
30.77
|
29.81
|
Energy yield (%) ***
|
NA
|
65.15
|
61.05
|
60.31
|
55.59
|
56.36
|
68.26
|
64
|
57.75
|
66.51
|
58.11
|
53.63
|
Hydrochar yield (%) ****
|
NA
|
54.27
|
51.12
|
47.66
|
42.3
|
40.57
|
59.54
|
53.99
|
41.44
|
50.96
|
40.47
|
38.56
|
The atomic ratios H/C and O/C of lavandin and hydrochars are presented in Fig. 2, as Van Krevelen diagram, in order to compare biomass and hydrochars with fossil fuels. Lavandin sample showed worse quality than peat, but its hydrochars showed wide range quality between bituminous coal to peat. Pathways for dehydration, and decarboxylation are also illustrated. From Fig. 2, dehydration and decarboxylation were the two main biomass transformations during HTC, rather than demethylation that should depict a decrease of H/C atomic ratio with an increase of O/C atomic ratio [51]. Hydrochars presenting properties closer to bituminous coal and lignite when the ones obtained for HTC conditions with severity factors higher than 6, indicating that high temperature and high retention time increased the coalification process of biomass.
SEM and ATR-FTIR
The surface morphologies of raw lavandin and its corresponding hydrochars are visualized in Figs. 3 and 4. Because of using the blender to crush the lavandin without the HTC process, broken branches and broken flowers can be seen from Fig. 3 (a). In Fig. 4 (a), raw lavandin showed fibrous structure with a smooth surface which is the typical structure of herbaceous biomass [52].
After the HTC process, each hydrochar had different levels of decomposition. The flower part completely disappeared after the HTC process. Corresponding to different experimental conditions, a great quantity of large-sized fibrous structures in LH1-220, LH2-220, LH4-220, LH4-180 and LH4-200 were preserved (Fig. 3 (b, c, d, g, h)), whose severity factors were less than 6. Further hydrolysis of cellulose or hemicellulose and partial hydrolysis of lignin, with the increase in severity factor, LH8-220, LH20-220, LH4-240, and LH20-200 samples were dominated by heterogeneous fragments so that only a few fibrous structures were observable as shown in Fig. 3 (e, f, i, g) [53]. When the severity factors were larger than 7, which represented the hydrochars LH20-240 and LH20-260 (Fig. 3 (k, l)), the fibrous structures were completely decomposed and disappeared, the heterogeneous fragments were reorganized into coal-like carbon particles, and it showed melted and agglomerated structures.
The fibrous structures in hydrochars are shown in Fig. 4 (b-l). After the HTC process, the surface of fibrous structure became rough, even the LH4-180, which had the lowest severity factor, showed some wrinkles and cracks on the surface and the original shape of the fibrous structure has been preserved [54, 55]. After that, samples showed different degrees of decomposition depending on the experimental conditions. Overall, with the gradual increase in severity factors, the surface tended to be rough and decomposed step by step, the tubular structure as Fig. 4 (e, f) and skeleton structure as Fig. 4 (h) inside were revealed and further decomposed. Another manifestation of fibrous structure shrinking was that its diameter was greatly reduced as shown in Fig. 4 (i, j). As shown in Figs. 3 (k, l), when severity factors were 7.2 and 7.9, respectively, the reaction conditions were sufficiently severe. It caused that the fibrous structure was exhausted and replaced by dense particles after hydrothermal carbonization.
The particle size of each sample was measured and plotted as a distribution histogram shown in Fig. 5. Lavandin was mainly dominated by particles larger than 100 µm. Because of the degradation during the HTC reaction, a large number of particles smaller than 100 µm appeared in the hydrochars, and this proportion showed a trend of first increasing and then decreasing with the increase in severity factor. This trend can also be seen from the comparison of SEM images. As the HTC reaction progresses, large particles decomposed into smaller particles, and then small particles reorganized and aggregated into larger particles. Finally, when the severity factor was larger than 7, corresponding to samples LH20-240 and LH20-260, the particle size distribution was similar. A higher temperature could lead to a more uniform average diameter and uniform size distribution [56].
Comparative ATR-FTIR spectra of the raw lavandin and its corresponding hydrochars are shown in Fig. 6. The band at 3600–3000 cm− 1 is attributed to the O-H stretching vibration in the hydroxyl or carboxyl groups [57]. When SF increased, especially when higher than 7, the intensity of the O-H stretching band decreased significantly, which indicates the results of dehydration and decarboxylation reactions during the HTC. The peaks at 2930 and 2850 cm− 1 are ascribed to C-H stretching vibration in aromatic and aliphatic structure in lavandin and hydrochars[58]. The characteristic C = O stretching of carboxylic acid groups in the hemicellulose appeared approximately at 1730 cm− 1 in the spectra of raw lavandin. It disappeared in its corresponding hydrochars due to decomposition of hemicellulose [59, 60]. The peak at around 1700 cm− 1 in all hydrochar samples is ascribed to the C = O groups in cellulose and lignin. When the temperature increased, relative intensification in this peak occurred due to the disappearance of the band at 1730 cm− 1 corresponding to the C = O bond in hemicellulose. The intensification points out that the decomposition of hemicellulose occurred at lower temperature compared with cellulose and lignin [61, 57]. C = C stretching vibration is seen at 1600 and 1500 cm− 1 and C-H bending vibration is seen at 1447 cm− 1, indicating the presence of mononuclear aromatic structures in L and all hydrochars. The aliphatic ether C-O stretching vibration is observable at 1160, 1100, 1060 and 1030 cm− 1 [55, 62]. Although the exact mechanism of the HTC reactions is still being elucidated, it can still be inferred that dehydration, decarboxylation and hydrolysis reactions likely occured during the HTC process.
Combustion behavior and thermal properties of lavandin and hydrochars
Thermodegradation of lavandin sample and hydrochars was performed under constant air flow (150 mL min− 1) and constant heating rate (10°C min− 1), in order to evaluate the combustion behavior of the samples. TGA and DTG curves for lavandin (Fig. 7 (a)) and its hydrochars obtained after a short retention time (from 1 hour to 4 hours) (Fig. 7 (b) to (d)), or a low temperature (below 220°C) and a relatively short time (4 hours) (Fig. 7 (g and h)) depicted the same trend: at around 100°C water evaporation and sublimation of low molecular weight component occurred, followed by a first degradation stage between 200°C and 400°C attributed to devolatilization and combustion, and finally a second degradation stage between 400°C and 650°C attributed to degradation products (char) combustion [47]. Ornaghi et al. [63] demonstrated that during thermodegradation of lignocellulosic fibers, water, extractives, hemicellulose and cellulose were totally degraded at temperature above 400°C, lignin was totally degraded at temperature above 550°C. Degradation of lignin at higher temperature than cellulose and hemicellulose was attributed to its complex, cross-linked three dimensional aromatic polymer structure, compared to the linear structure of cellulose and branched and amorphous polymer structure of hemicellulose. Chars of hemicellulose and cellulose started to degrade at 300°C while lignin char started to degrade at 400°C. Considering the experimental conditions used to obtain hydrochars LH1-220, LH2-220, LH4-220, LH4-180, LH4-200, TG/DTG curves presenting a two stage degradation process confirmed that extractives, cellulose, hemicellulose and lignin were partially degraded during HTC process. The calculated severity factors for these hydrochars were very narrow, ranging from 4.74 to 5.91. Hence, the time-temperature combinations chosen for these samples provided approximatively the same kind of hydrochar, with comparable thermal properties. The calculated kinetic parameters are presented in Table 3. First order kinetic model gave the best correlation coefficients ranging from 0.93 to 0.99 for lavandin sample (L) and all its hydrochars. Considering the two thermodegradation stages appearing on TG/DTG curves for L, LH1-220 to LH20-220, LH4-180 and LH4-200, kinetic parameters were determined for each one of them (Table 3). Activation energy and pre-exponential factor for lavandin sample (L) were lower for both stages than LH1-220, LH2-220, LH4-220 hydrochars, indicating that L sample was more reactive to ignite than hydrochars obtained at 220°C. When retention time was increased to 8h or 20h and temperature was kept constant (220°C), activation energy and pre-exponential factor were lower than for lavandin sample, indicating that hydrochars were more reactive to ignite than lavandin. When retention time was kept constant (4h) and temperatures were lower than 220°C, the activation energy and pre-exponential factor were comparable to the ones obtained for LH1-220 to LH4-220. Regarding hydrochars obtained at temperatures larger than 220°C and/or retention time larger than or equal to 4h, distinction of several degradation stages is not as obvious as previously. TG/DTG curves depicted a single stage thermal decomposition covering a large temperature range from 200°C to 800°C, probably indicating that extractives, hemicellulose and cellulose were already totally or partially degraded during HTC process. TG/DTG curves may correspond (i) to a lesser extent to the thermal degradation of the remaining extractives, cellulose and hemicellulose and (ii) in a larger extent to chars degradation. Activation energy was around two times lower than for others hydrochars, indicating a good ignition, and pre-exponential factor was very low. In general, the pre-exponential factor gives values indicating the fraction of reactant molecules possessing enough kinetic energy to react. In the particular HTC process, its values are located in a wide range as the thermal degradations occur in solid phase. As shown in Table 3, the A values are in the range 185.4–0.001 min− 1. It was noticeable that LH4-240, LH20-240, LH20-260 and LH20-220 hydrochars have the lowest values of A factor, signature of a surface reaction. The severity factor values for these hydrochars were higher than the previous ones (superior to 6), and the combination of time and temperature gave hydrochars with satisfactory thermal properties.
Table 3
Combustion kinetic parameters (E, A, n, R2) of Lavandin (L) and its hydrochars (LH) obtained at different temperature and different retention time. Ignition temperature (Ti), maximum combustion rate temperature (Tm), and burnout temperature (Tb); S: combustibility index\(= \frac{{\left(\frac{dw}{dt}\right)}_{max}{\left(\frac{dw}{dt}\right)}_{mean}}{{T}_{b} \times { T}_{i}^{2}}\), with \({\left(\frac{dw}{dt}\right)}_{max}\) and \({\left(\frac{dw}{dt}\right)}_{mean}\) the maximum and mean weight loss rate (%/min), respectively[64]
|
SF
|
Temperature range (°C)
|
E (kJ/mol)
|
A (min− 1)
|
n
|
R2
|
Ti (°C)
|
Tb(°C)
|
S (10− 8)
(%2min−2 °C− 3)
|
L
|
NA
|
160–400
|
26.43
|
11.5
|
1
|
0.98
|
230
|
505
|
19.8
|
|
|
400–540
|
38.46
|
100.7
|
1
|
0.94
|
|
|
|
LH1-220
|
5.31
|
219–386
|
34.15
|
31.0
|
1
|
0.97
|
270
|
553
|
13.5
|
|
|
386–590
|
43.66
|
165
|
1
|
0.95
|
|
|
|
LH2-220
|
5.61
|
225–389
|
33.61
|
24.3
|
1
|
0.97
|
270
|
570
|
12.1
|
|
|
389–595
|
43.37
|
132.8
|
1
|
0.94
|
|
|
|
LH4-220
|
5.91
|
231–384
|
30.61
|
10.1
|
1
|
0.97
|
270
|
582
|
11.0
|
|
|
384–610
|
45.87
|
185.4
|
1
|
0.95
|
|
|
|
LH8-220
|
6.21
|
235–373
|
24.15
|
1.8
|
1
|
0.99
|
312
|
646
|
5.25
|
|
|
373–685
|
39.22
|
36.3
|
1
|
0.95
|
|
|
|
LH20-220
|
6.61
|
235–292
|
18.61
|
0.4
|
1
|
0.99
|
330
|
662
|
4.43
|
|
|
292–692
|
32.87
|
10
|
1
|
0.95
|
|
|
|
LH4-180
|
4.74
|
212–379
|
33.16
|
32.3
|
1
|
0.98
|
260
|
615
|
11.9
|
|
|
379–664
|
30.16
|
12.3
|
1
|
0.98
|
|
|
|
LH4-200
|
5.32
|
212–379
|
34.23
|
34.3
|
1
|
0.97
|
270
|
615
|
10.5
|
|
|
379–664
|
34.22
|
24
|
1
|
0.93
|
|
|
|
LH4-240
|
6.50
|
212–782
|
13.87
|
0.001
|
1
|
0.99
|
280
|
821
|
3.06
|
LH20-200
|
6.02
|
212–517
|
27.3
|
8.7
|
1
|
0.99
|
260
|
880
|
3.25
|
LH20-240
|
7.20
|
212–940
|
15.52
|
0.2
|
1
|
0.99
|
300
|
950
|
2.38
|
LH20-260
|
7.79
|
212–782
|
17.24
|
0.3
|
1
|
0.98
|
300
|
851
|
2.96
|
Ti and Tb (Table 3) obtained for lavandin sample (230°C and 505°C, respectively) were lower than Ti and Tb obtained for hydrochars (270°C to 330°C and 553 to 950°C, respectively), indicating that hydrochars were better biofuels than raw lavandin. Low Ti values is governed by the amount of volatile matter in the hydrochar: as high the volatile matter content is, as low Ti is. High ignition temperature indicated that hydrochars used as solid biofuels could be safely stored at ambient temperature and transported, without risk of auto-ignition. The highest Ti values corresponded to hydrochars obtained for a retention time superior to 8h at moderated HTC-temperature (180–220°C), or at high HTC-temperatures (240–260°C). Burnout temperature increased as the retention time and/or the HTC-temperature increased. The highest Tb values were obtained at the highest HTC-temperature and the longest retention time (240–260°C; 20h). The increase of burnout temperature is correlated with the increase of fixed carbon measured in hydrochars. Longer the retention time during HTC (and/or higher temperature) is, more thermally stable the carbon components in hydrochars are.
Combustibility index is presented as a function of severity factor used during HTC experiments in Fig. 8. From this figure, differences in combustibility index clearly appeared as two groups of datasets: the first one corresponded to the combustibility index for hydrochars obtained at SF below 6, and the second one corresponded to the combustibility index for hydrochars obtained at SF above 6. For SF < 6, the combustibility index was two to three-fold superior to the combustibility index obtained for SF above 6, indicating that these hydrochars were more suitable for combustion purposes.