Article pubs.acs.org/EF
Structural Characterization of Large Polycyclic Aromatic Hydrocarbons. Part 2: Solvent-Separated Fractions of Coal Tar Pitch and Naphthalene-Derived Pitch Valentina Gargiulo,*,† Barbara Apicella,† Fernando Stanzione,† Antonio Tregrossi,† Marcos Millan,‡ Anna Ciajolo,† and Carmela Russo† †
Istituto di Ricerche sulla Combustione, Consiglio Nazionale delle Ricerche (CNR), Piazzale Vincenzo Tecchio 80, 80125 Napoli, Italy ‡ Department of Chemical Engineering, Imperial College London, London SW7 2AZ, United Kingdom ABSTRACT: Complex polycyclic aromatic hydrocarbon (PAH) mixtures separated from a coal tar pitch (CP) and naphthalene pitch (NP) by sequential extraction with heptane and toluene were characterized in detail by applying a multiarray analytical approach. Gas chromatography−mass spectrometry (GC-MS), size exclusion chromatography (SEC), laser desorption ionization-time-of-flight mass spectrometry (LDI-TOFMS), and thermogravimetry (TG) were used to relate the volatility and coking yield of pitch components to their solubility and molecular weight distribution. Spectroscopic analysis, including infrared (IR), ultraviolet−visible (UV−vis), and fluorescence spectroscopy, proved to be useful for measuring specific features of aromatic systems, such as the aromatic content, degree of aliphatic substitution, and size distribution of PAHs of different molecular weights. In particular, it has been shown that the spectroscopic analysis is an essential tool for characterizing very large PAH systems concentrated in the pitch toluene-insoluble fraction. This fraction constitutes a case study of very large, structurally different aromatic compounds, and it is the pitch fraction more relevant for practical applications because of its higher coking tendency and peculiar optical properties.
1. INTRODUCTION Fossil fuel-derived and synthesized pitches are important sources of polycyclic aromatic hydrocarbons (PAHs), which can be employed in manufacturing various carbon-based materials such as graphite electrodes, carbon anodes, carbon fibers, and composite fibers. Details on the PAH structures, their interactions, and their assembly mechanism are important as they determine the properties of carbon materials obtained following heat treatment. Moreover, pitches constitute a case study of readily available mixtures of large PAHs that can be traced to the nanosized sp2 domains featuring ordered and disordered graphenic materials. The complexity and broad molecular weight (MW) distributions of pitches require a multiarray analytical approach to elucidate their quality and properties. This can be performed on the raw pitches but works much better on their molecular weight subfractions. The fractionation of very complex carbonaceous pitch samples has been achieved by different methods including solvent solubility,1−3 column chromatography on silica,4 preparative or analytical size exclusion chromatography (SEC),5 thin layer chromatography (TLC),6 and ultrafiltration membranes.7 For simpler pitches, such as pyrene pitches, supercritical fluid extraction (SFE) in N-methyl-2-pyrrolidinone (NMP)/toluene mixtures has also been applied.8 However, the limitation of chromatographic methods and ultrafiltration membranes was the small quantity of sample recovered in each fraction, and in the case of SEC and SFE the solvent in which the recovered fractions were dissolved (NMP). On the other hand, solvent extraction techniques are amenable to be © XXXX American Chemical Society
applied at larger scales. Specifically, extraction/solubilization using quinoline (standard ASTM D2318 or ASTM D7280 test methods) or toluene (standard ASTM D4312 or ASTM D4072 test methods) has become the main practical method to assess pitch quality9 in practical industrial applications. The relative proportions of quinoline-insoluble and toluene-insoluble fractions determine the pitches softening and coking properties, which in turn are important for their applicability in various manufacturing processes. Solvent extraction has also been applied to remove the more soluble and typically also more volatile fraction, leaving behind a more insoluble fraction, prone to form mesophases and useful for industrial applications.10,11 However, solubility cannot be used as a sole criterion to predict mesophase properties because these are also related to the nature of the precursor and the mesophase preparation procedure.12,13 For analytical purposes, successive extractions with heptane and toluene have been used to separate the lighter pitch components in order to allow the characterization of the higher MW aromatic species concentrated in the tolueneinsoluble fraction.14,15 Following our recent work (10.1021/acs.energyfuels.5b01327),2 the characteristics of structurally different PAH systems constituting a coal tar pitch (CP) and a synthetic naphthalene pitch (NP) have been further investigated with the analysis of the pitch solvent-separated fractions. The fractions separated by successive extraction with heptane and toluene Received: November 2, 2015 Revised: February 19, 2016
A
DOI: 10.1021/acs.energyfuels.5b02576 Energy Fuels XXXX, XXX, XXX−XXX
Article
Energy & Fuels
dispersions were compressed at 10 Ton for 10 min into thin disks having a mean thickness value of 0.035 cm. Thermogravimetry (TG) analysis was performed on a PerkinElmer Pyris 1 thermogravimetric analyzer. The samples were heated from 50 °C up to 750 °C at a rate of 10 °C min−1. Experiments were performed in both inert (N2, 40 mL min−1) and oxidative (air, 30 mL min−1) atmospheres. Ultraviolet−visible (UV−vis) spectra of the parent pitches and the solvent-separated fractions dissolved/suspended in NMP were measured from 260 to 1100 nm on a HP 8453 diode array spectrophotometer using 1 cm path-length quartz cuvettes. The mass-specific absorption coefficients were measured for solutions/ suspensions with a concentration of 5 mg L−1. Fluorescence spectra were acquired on a PerkinElmer LS-50 spectrofluorometer. The wavelength accuracy was ±1.0 nm, and the wavelength reproducibility was ±0.5 nm. The fluorescence measurements were performed on samples highly diluted in NMP (from 0.5 to 2 mg L−1) to avoid concentration quenching and other phenomena that can distort the spectrum and affect fluorescence intensity and quantum efficiency. Synchronous fluorescence emission spectra were measured by applying simultaneous scanning of the excitation and emission wavelengths keeping their difference (Δλ = 10 nm) constant. The quantum efficiencies of the samples were evaluated with a comparison to 9,10-diphenylanthracene,27 which is a standard molecule assumed to have a unitary quantum yield.29
were characterized using the chemical, thermal, and spectroscopic techniques previously employed to analyze the raw pitches.2 In respect to many previous works regarding the characterization of coal tar pitch (e.g., the work of Zander14) and naphthalene pitch (e.g., the work of Mochida16), structural features as the hydrogen quality and the aromatic character of smaller and larger PAH systems featuring CP and NP have been inferred by spectroscopic analysis.
2. EXPERIMENTAL SECTION The CP sample obtained from high-temperature coking of a British coal2,3 has a hydrogen/carbon (H/C) atomic ratio of 0.54 and the following elemental composition: C = 91.4 wt %, H = 4.1 wt %, N = 1.32 wt %, S = 1.76 wt %. The NP sample is an AR-mesophase pitch provided by Mitsubishi Gas-Chemical Company with H/C atomic ratio equal to 0.59 and very low amounts of heteroatoms (0.23 wt % of sulfur and below 0.1 wt % of nitrogen and oxygen).16 The sequential extraction of CP and NP with heptane and toluene had the following results, respectively: 14.4 and 0.7 wt % heptane solubles (HS), 44.4 and 30.1 wt % toluene solubles (TS), and 41.2 and 69.2 wt % toluene insolubles (TI). In spite of their similar H/C ratios,2 NP was much less soluble than CP. This is the first important indication of their different chemical properties/structures, which in turn determine different physical properties such as softening points and volatilization. It is worth to note that, unlike heavy oils,17,18 the TS fractions of CP and NP were completely soluble in NMP. This can be attributed to the larger aromatic content of pitch-derived than oil-derived fractions and the great effectiveness of NMP in solubilizing aromatic-rich mixtures.19 Gas chromatography−mass spectrometry (GC-MS) analysis was carried out on an Agilent HP6890/HP5975. The gas chromatograph was equipped with a DB-5MS capillary column (60 m × 0.25 mm i.d., 0.25 μm film thickness). Helium was used as carrier gas at a constant flow of 1.0 mL min−1. For the analysis of CP-HS and CP-TS the oven temperature was programmed to increase from 40 °C (4 min) to 300 °C at a heating rate of 5 °C min−1 and then held for 15 min. For the analysis of NP soluble fractions, which are heavier than those of CP, the oven temperature was increased to a final value of 320 °C. A sample volume of 1 μL was injected (sample concentration ranging between 1000 to 2000 mg L−1). The mass spectrometer operated in electron ionization mode, and m/z was scanned from 50 to 600. The concentration of each PAH is quantified using response factors calculated from the analysis of a standard PAH mixture (MW up to 300 Da supplied by Supelco EPA 525 PAH mix A) of known concentrations. The SEC analysis of pitch samples and solvent-separated fractions was carried out on a HPLC system HP1050 series by elution with NMP on a PL-gel polystyrene−polydivinylbenzene individual-pore column (Polymer Laboratories Ltd., UK, part no. PL1110-6525) for the MW determination in the 100−50 000 Da range.2,20−22 The online detection of species eluted from the SEC column used a HP1050 UV− visible diode array detector that measured the absorbance signal at fixed absorption wavelengths (350 and 500 nm). The laser desorption ionization-time-of-flight mass spectrometry (LDI-TOFMS) spectra were recorded in positive reflectron mode on a Voyager DE STR Pro instrument (Applied Biosystems, Framingham, MA). Matrices were not added as all of the investigated samples are able to absorb the laser beam (λ = 337 nm) acting as a self-matrix.23−26 More details on the sample preparation, instrument setups, and data analysis have been given elsewhere.2,24 Fourier transform infrared (FTIR) spectra in the 3400−600 cm−1 range were acquired in transmittance mode using a Nicolet iS10 spectrophotometer. The quantitative FTIR analysis of aliphatic and aromatic hydrogen27 was performed on solid dispersions prepared by mixing and grinding the samples (0.25−0.5 wt %) with KBr. The KBr
3. RESULTS AND DISCUSSION 3.1. GC-MS Analysis of Soluble (HS and TS) Pitch Fractions. The GC-MS profiles of CP-HS and CP-TS, reported in Figure 1, showed a sequence of PAHs from two
Figure 1. GC-MS profiles of CP-HS (upper panel) and CP-TS (lower panel).
to seven rings, as previously found in a coal tar pitch30 as well as in the fraction distilled from coal tar (anthracene oil).31 These PAHs can be grouped mainly into four categories: (i) alternant pericondensed, such as pyrene and benzo[a]pyrene; (ii) nonalternant pericondensed, such as fluoranthene and benzo[b]fluoranthene; (iii) catacondensed, such as phenanthrene, anthracene, benzo[a]anthracene, and chrysene; and (iv) derivatives, such as acenaphthene and fluorene in which methylene groups are inserted in the PAH structure.32,33 The PAHs observed by quantitative GC-MS analysis accounted for approximately 40 wt % of the CP-HS. Traces of methyl-, oxygen- (furans), nitrogen- (carbazoles), and sulfur(tiophenes) substituted PAHs were also detected. The TS B
DOI: 10.1021/acs.energyfuels.5b02576 Energy Fuels XXXX, XXX, XXX−XXX
Article
Energy & Fuels
Figure 2. TG profiles of the CP and NP solvent-separated fractions in inert (N2, left panels) and oxidative (air) (right panels) atmospheres.
fraction (lower panel of Figure 1) showed a similar PAH distribution as CP-HS although enriched in four- to seven-ring PAHs. These PAHs, however, account for less than 20 wt % of the CP-TS sample, indicating that toluene effectively extracted species with greater MW than the GC analysis limit (m/z 350− 400).33 The presence of the same medium-size (four- to sevenring) PAHs in the HS and TS fractions may reflect a limited solubility or the fact that the solubility of a species in a complex mixture, such as pitch, depends not only on the MW and chemical structure, but also on the matrix in which it is imbedded. The occurrence of oligomeric structures entangling smaller PAHs in their interstices and limiting their solubilization was envisioned by Zander in his thorough work on pitch characterization.14 In addition to the NP minimal solubility (negligible in heptane and relatively low in toluene), typical of anisotropic pitches, the GC-MS analysis of NP-HS and NP-TS was almost completely ineffective. Only peaks relative to alkyl-substituted PAHs (e.g., methyl-substituted benzo[ghi]perylenes) could be seen in the GC-MS chromatogram of NP-HS. This clearly demonstrates that even the soluble components of NP have a very high MW. This is better shown by the following MW analysis, which also demonstrates that the presence of light PAHs possibly entrapped in the toluene insolubles could be excluded. 3.2. TG Analysis. The TG profiles of the CP and NP solvent-separated fractions measured in inert and air atmospheres are reported in Figure 2. These profiles enabled volatilization ranges, coking yields, and oxidation reactivities to be determined. The TG curves of raw pitches are also reported in Figure 2 to show the contribution of each fraction to the thermal properties of parent pitches. Overall, TG analysis in an inert environment showed that the HS and TS fractions, which are richer in small- and mediumsize PAHs (Section 3.1), are the more volatile and they are responsible for the weight loss of parent pitches, in particular for the CP weight loss between 100 and 400 °C. The volatilization of CP-TS occurred at higher temperatures and for a lesser extent compared to CP-HS. These results confirmed
the enrichment of heavier components in the TS sample indicated by GC-MS analysis (Section 3.1). Considering the almost negligible amount of NP-HS (Section 2) and the lack of volatiles shown by the NP-TI (Figure 2), the weight loss of NP (25 wt %) observed between 400 and 500 °C range in an inert atmosphere can be exclusively ascribed to NP-TS. The weight loss of NP-TS at high temperatures (>400 °C) is consistent with a combination of the vaporization of species having higher MW than the upper limit (m/z 350−400) of the GC-MS analysis and the occurrence of decomposition reactions with subsequent volatilization of the products. CP-TI and NP-TI present similar TG profiles with almost negligible weight losses in N2, illustrating the low volatility and the greater coking tendency typical of toluene insolubles.34 This thermal behavior reflects the heaviness of TI components, and it is due to the occurrence of condensation/polymerization reactions within the TI aromatic matrix upon heating.1,35 The heat treatment in an oxidative atmosphere appeared to enhance the coking yield of the HS and TS fractions as indicated by the weight loss reduction (right part of Figure 2) with respect to TG experiments carried out in an inert atmosphere. A small but significant weight gain for the TI fractions was clearly observed (especially for NP-TI), demonstrating their high propensity toward oxygen addition during the heat treatment. This behavior could favor the subsequent complete burnoff at a lower temperature (around 650 °C) than to HS and TS fractions. However, a better contact between oxygen and the porous surface and granular texture of TI fractions can also explain their higher oxidation rate with respect to the HS and TS fractions, which show a shiny, tarry surface less permeable to oxygen. 3.3. SEC. The MW distribution of pitch solvent-separated fractions has been estimated by SEC coupled with UV−vis detection, a method frequently used for heavy samples derived from coal, biomass, and petroleum.36 A calibration of the SEC system with polystyrene and PAH standards22 has allowed the transformation of SEC profiles into MW distributions. The area-normalized SEC profiles of the solvent-separated fractions multiplied by their weight percentages (Section 2) are reported C
DOI: 10.1021/acs.energyfuels.5b02576 Energy Fuels XXXX, XXX, XXX−XXX
Article
Energy & Fuels in Figure 3 to show their contributions to the MW distribution of CP and NP.
complete overlapping of CP-HS profile with the lower MW tail of the CP-TS profile (Figure 3) indicates that the same light species are solubilized in heptane and toluene, in agreement with the GC-MS analysis (Section 3.1 and Figure 1). For both CP and NP, a significant overlapping also occurred between the heavy end of the TS fraction and the light tail of the TI fraction. This overlapping extended across a wider MW range in the case of CP-TI and CP-TS than the corresponding NP fractions, and indeed, the lighter tail of CP-TI appeared partly superimposed even with CP-HS (Figure 3). As previously mentioned (Section 3.1), the entanglement of lighter and medium-size PAHs in the CP-TS and CP-TI matrices14 could be considered responsible for this overlapping. However, the nature of the solvent fractionation, in which not only MW but also chemical structures have an impact on the composition of the different fractions, can also contribute. 3.4. LDI-TOFMS. The mass spectra of CP and NP are reported in the upper part of Figure 4, and the mass spectra of their corresponding TS and TI fractions are reported in the middle and lower parts of Figure 4, respectively. Mass spectrometric analysis of the HS fractions was not carried out due to the almost negligible amount of NP-HS and the high volatility of CP-HS, which was thoroughly characterized by GC-MS analysis (Section 3.1 and Figure 1). As reported in the previous work (10.1021/acs.energyfuels.5b01327),2 the LDI-TOFMS analyses of CP and NP exhibited a continuous mass sequence from m/z 150 to approximately 1000. The mass spectra of the TS fractions were centered in the m/z 200−800 range and extended up to approximately m/z 1000, reproducing the spectra of parent
Figure 3. SEC profiles of the solvent-separated fractions of CP (upper panel) and NP (lower panel).
The MW maximum progressively shifts toward higher values from HS to TS and TI fractions for both CP and NP. The
Figure 4. LDI-TOFMS (reflectron configuration) spectra of CP, CP-TS, and CP-TI (left column) and NP, NP-TS, and NP-TI (right column). D
DOI: 10.1021/acs.energyfuels.5b02576 Energy Fuels XXXX, XXX, XXX−XXX
Article
Energy & Fuels
Figure 5. LDI-TOFMS (reflectron configuration) spectra of CP-TI and NP-TI in the m/z 100−2000 range.
pitch components and calibration samples (polystyrenes and PAH standards) used for the MW attributions, however several studies by Herod and Kandiyoti (e.g., the work of Karaca and references therein)20 led to the conclusion that PS calibrations were reliable at least up to MW of around 3000 Da. In the case of CP-TI, the Mn value evaluated by SEC appears to be anomalously high (1904 Da) compared to that evaluated by LDI-TOFMS (789 Da). In CP-TI, this could result from the presence of high MW carbonaceous species that were not desorbed and ionized in the LDI-TOFMS. In alternative, such high-MW carbonaceous species have been fragmented in the flight tube, causing their appearance at lower m/z in the reflectron configuration used in this work.38 In fact, both the mass spectrometric and SEC profiles (Figure 5 and Figure 3, respectively) of CP-TI exhibited a tail in the lower end, indicating the presence of light components ascribable to fragmentation and/or entanglment with the high-MW components. Conversely, such low-MW species were not present neither in the SEC profile (Figure 3) or in the mass spectrum of NP-TI (Figure 5). Only a group of peaks at approximately m/z 200, with a very low intensity and good separation from the main mass distribution that peaked around m/z 600−700, was visualized in the mass spectra of NP-TI and parent NP, indicating the occurrence of some prompt fragmentation in the ion source.38 Notably, the mass spectrometric analysis shows that the Mn values of both TI fractions are approximately twice the Mn values of TS fractions suggesting that the TI fractions could be change composed mainly of oligomers of PAHs contained in the TS fractions. To this regard, intramolecular reactions of these oligomers could also be considered responsible for the higher TI coking yield (around 80−90 wt %) (Section 3.2, Figure 2). This important feature will be deepened in a future work.
pitch fairly well (Figure 4). Compared to the TS fractions, the mass spectra of TI fractions were clearly shifted to higher MW, particularly in the case of NP-TI. This confirms that the insolubility of pitch components is mainly associated with higher MW species. The extension of the mass range of the TI fractions toward higher values is better shown in Figure 5, where the mass spectra of CP-TI and NP-TI up to m/z 2000 are contrasted. In Figure 5 a longer tail at m/z higher than 1000 was noticeable, which was not clearly visible in the spectra of raw pitches (Figure 4). Considering the important role of the heavier components in the physical, chemical, and thermal properties of pitches, the LDI-TOFMS analysis of the raw pitch could be misleading, producing distributions and number-average MW (Mn) biased toward lower molecular masses. Interestingly, in the previous work (10.1021/acs.energyfuels.5b01327) the Mn values evaluated from the mass spectrometric profiles of pitch samples were lower than those evaluated by SEC analysis.2 This discrepancy was particularly evident for CP, which was characterized by the presence of species distributed over a very large mass range. The Mn values of raw pitches and solvent-separated fractions are reported in Table 1. It can be observed that the Mn values Table 1. Mn Values of CP, NP, and Their Solvent-Separated Fractions Evaluated by SEC and LDI-TOFMS in the 150− 4000 Da Rangea Mn CP CP-TS CP-TI NP NP-TS NP-TI a
LDI-TOFMS
SEC
417 455 789 691 529 960
796 524 1904 824 561 1106
3.5. SPECTROSCOPIC ANALYSIS Spectroscopic methods, including FTIR and UV−vis absorption and fluorescence emission, have been applied to give insight into the different functionalities and, in particular, into the different aromatic characteristics of the pitch solventseparated fractions. Due to the very low amount of NP-HS fraction, its spectroscopic properties have been not evaluated as its contribution to the optical properties of the parent NP is considered almost negligible. 3.5.1. FTIR Spectroscopy. The FTIR spectra of CP and NP, presented in Figure 6 with the spectra of the more abundant TS and TI fractions, provided more details on the functional groups and chemical features of pitches.1,32,39,40 All
The statistical error is approximately 15%.
evaluated by LDI-TOFMS, especially for CP, are relatively low and do not account for the very high MW of CP-TI measured by MS. This confirms that mass spectrometry of mixtures of light and heavy species underestimates the contribution of higher MW species, probably because of their lower detection efficiency and/or higher propensity to fragment in comparison to low-MW species.37 The Mn values observed by SEC for TS and TI fractions are noticeably higher than those measured by LDI-TOFMS. This can be merely due to some structural differences between the E
DOI: 10.1021/acs.energyfuels.5b02576 Energy Fuels XXXX, XXX, XXX−XXX
Article
Energy & Fuels
Figure 6. FTIR absorption spectra of CP, CP-TS, CP-TI (left part) and NP, NP-TS, NP-TI (right part).
spectra show similar band groups, differing from each other in their relative intensity in the high-frequency region (3100− 2800 cm−1), due to C−H stretching vibrations; in the mediumfrequency region (1700−1000 cm−1), typical of complex aromatic defective networks;27,41,42 and in the low-frequency region (1000−600 cm−1) due to aromatic C−H out-of-plane (OPLA) bending modes. In the CP spectrum the intensity of the aromatic C−H stretching peak (at approximately 3050 cm−1) is higher compared to that of the aliphatic C−H stretching signals (peaks around 2925 cm−1). The opposite is found in the NP spectrum. These features are preserved in the relative TS and TI fraction spectra. In the stretching region of the samples reported in Figure 6, the contribution of aliphatic hydrogen was noticeably higher in the TS fractions than in the TI fractions for both NP and CP. The other region differentiating CP and NP was the 900−700 cm−1 wavenumber region where OPLA bending modes occurred: solo (890−870 cm−1), duo (850 and 810 cm−1), and trio/quatro aromatic C−H bonds (790−720 cm−1).27,41,43 The trio/quatro hydrogen peaks of CP and its fractions were the most intense in the OPLA spectral region, as previously found in coal tar pitch.1 This indicates the apparent predominance of PAHs with ortho-fused end rings34 in the CP fractions, which justifies the relevant solubility of CP in heptane and toluene (Section 2).14 An almost uniform distribution of the OPLA peaks was generally observed for NP fractions, suggesting that highly substituted aromatics are present in both TS and TI fractions. In fact, the NP-TI spectrum did not show significant differences compared to that of NP-TS, except for a lower relative abundance of aliphatic signals. Conversely, the lower intensity of OPLA signals, the higher intensity of the aromatic CC stretching peak (1600 cm−1) and the broad absorption in the 1300−1100 cm−1 region observed in the CP-TI spectrum confirmed the significant contribution of carbon-rich species with complex aromatic networks and the higher molecular masses observed by SEC analysis (Table 1). The quantitative FTIR method set up in a previous work20 and used in the previously published part (10.1021/ acs.energyfuels.5b01327) of this study for the parent CP and NP analysis2 has also been applied to determine the aliphatic (Hal) and aromatic hydrogen (Har) in the solvent-separated fractions. Table 2 reports the aliphatic and aromatic hydrogen contents and the H/C ratios of the TS and TI fractions and of
Table 2. FTIR Parameters of CP, NP, and Their SolventSeparated Fractions aromatic hydrogen, Har (wt %) aliphatic hydrogen, Hal (wt %) hydrogen total (wt %) H/C (atomic) methyl carbon, CH3 (wt %) methylene carbon, CH2 (wt %) methyne carbon, CH (wt %) protonated aromatic carbon (wt %) sp3-bonded carbona (wt %) sp2-bonded carbonb (wt %)
CP
CP-TS
CP-TI
NP
NP-TS
NP-TI
3.16
3.58
2.07
1.87
1.69
1.85
0.86
1.06
0.55
2.42
2.99
2.13
4.01
4.64
2.62
4.29
4.68
3.97
0.50 0.92
0.58 1.02
0.32 0.62
0.54 1.38
0.59 1.87
0.50 1.32
2.92
3.97
1.79
10.63
12.36
8.94
1.66
1.67
1.20
3.65
5.57
3.67
37.91
42.97
24.77
22.38
20.27
22.18
5.50
6.66
3.61
15.66
19.80
13.93
90.48
88.70
93.77
80.06
75.52
82.10
a
sp3-bonded carbon is the sum of aliphatic (CH3 + CH2 + CH) carbon percentages. bsp2-bonded carbon is the total carbon subtracted of aliphatic carbon.
the parent pitches. Table 2 also shows the percentage of protonated aromatic carbon (hydrogen-bonded aromatic carbon) and sp3-bonded aliphatic carbon (CH, CH2, and CH3), which were directly evaluated from the quantitative FTIR analysis of aromatic and aliphatic hydrogen. The percentage of sp2-bonded aromatic carbon (calculated subtracting the total hydrogen and aliphatic carbon contributions from the total pitch mass) are also listed in Table 2. The hydrogen content of CP and its related fractions was mainly due to aromatic hydrogen. CP-TS exhibited a hydrogen content (H/C = 0.58) much higher than that of CP-TI (H/C = 0.32). The low H/C value for CP-TI and the appearance of a broad spectral feature in the 1300−1100 cm−1 range in the FTIR spectrum (left part of Figure 6) are typical of carbon-rich materials as soot.27 These carbonaceous species cannot be desorbed, ionized, and eventually detected by mass spectrometric analysis, and their presence may partially explain the lower MW values determined from LDI-TOFMS than those obtained from SEC analysis (Table 1). F
DOI: 10.1021/acs.energyfuels.5b02576 Energy Fuels XXXX, XXX, XXX−XXX
Article
Energy & Fuels
Figure 7. UV−vis absorption spectra of the solvent-separated fractions of CP (left panel) and NP (right panel).
Table 3. Mass-Specific Light Absorption Coefficients (cm2 g−1) of CP, NP, and Their Solvent-Separated Fractions
The hydrogen of NP and its related fractions is mainly aliphatic, as found also for petroleum pitch.44 Moreover, the decrease in the hydrogen content going from NP-TS (H/C = 0.59) to NP-TI (H/C= 0.50) was relatively low and resulted exclusively from the decrease in aliphatic hydrogen (from 2.99 to 2.13 wt %). Due to the peculiar spectral features of the CH2 symmetric stretching peak, the significant presence of CH2 groups in NP-TI and NP-TS (Table 2) can be attributed to aliphatic carbons bridging aromatic rings or included in naphthenic groups, as previously found for NP2. FTIR analysis has also been used to evaluate sp2-bonded carbon through the quantification of aromatic hydrogen and aliphatic hydrogen in the CH, CH2, and CH3 forms.2,27 It is worth noting that the TI fractions exhibited a higher percentage of sp2-bonded carbon (Table 2), consistent with their lower hydrogen content. More information on the nature of the sp2bonded carbon was obtained by the UV−vis and fluorescence spectroscopy, reported below. 3.5.2. UV−vis Spectroscopy. The profiles of the massspecific light absorption coefficients of the solvent-separated fractions multiplied by their weight percentages (Section 2) are reported in Figure 7 to show their contribution to CP and NP spectra (also reported). In the left panel of Figure 7, it can be seen that CP-HS and CP-TS were the main contributors to the UV−vis spectrum of CP up to approximately 500 nm, reproducing both the fine structure and the strongly decreasing intensity toward the visible region. These features are typical of a series of individual ortho-fused end ring PAHs30,45 from naphthalene to coronene, including PAHs that belong to either the so-called K-region PAHs or the phenacenes series.46,40 The analogous fine structure of CP-HS and CP-TS spectra confirms the presence of similar PAHs, as detected by GC-MS (Section 3.1). However, despite having a lower content of chromatographable PAHs, CP-TS showed a higher absorption coefficient (Table 3) than CP-HS in the whole spectrum (from 300 to 500 nm). This result indicates that the unidentified components of CP-TS have to be heavy aromatic species that exceed the GC mass limit (>350 Da).33 Notably, the CP absorption in the visible region was exclusively due to the TI fraction (Table 3). The CP-TI spectrum exhibited distinct features, with a broad featureless shape and a continuous smooth absorption decrease approaching the visible region, which is typical of complex carbon networks.47 This observation confirms the predominant contribution of carbon-rich species ascribable to impurities (soot, coke) derived from the CP production process, in agreement with FTIR, LDI-TOFMS, and SEC analyses.
CP CP-HS CP-TS CP-TI NP NP-TS NP-TI
300 nm
350 nm
400 nm
500 nm
600 nm
83800 65500 74600 90900 52500 89500 35600
46800 20100 31300 67300 38700 56100 29200
29600 6100 15600 50000 25000 33200 20600
13000 200 2800 29300 16600 19300 14800
6400 300−400 Da) to be analyzed by gas chromatography. Infrared spectroscopy analysis put in evidence the relative high concentration of aliphatic (mainly naphtenic groups) at the edges of the aromatic moieties. The high molecular weight of the NP components caused the very low volatilization and the high coking yield of NP which are typical of anisotropic pitches. Consistently with the polymerization route featuring NP production, the insolubles in toluene presented an average molecular weight that was about twice that of the toluene solubles. Most data on the insoluble (heavy) fractions of both CP and NP, which are the most relevant fractions for practical applications, was obtained through infrared, UV−vis and fluorescence spectroscopy allowing a better understanding of pitches at the molecular structure level. From the fundamental point of view this study contributes to give further information on the structure, configuration and relative optical properties of large PAHs not easily available and difficult to synthesize. As regards the relevance for practical applications, the deep characterization of solvent-separated pitch fractions gives information on specific pitch features like the volatility, the molecular weight distribution, and the aliphaticity/aromaticity of the carbon network. These characteristics are important for understanding the structure−property relationships, which in turn have effect on the pitch processing (thermal or catalytic) for advanced carbon materials fabrication. It has been shown that the solvent extraction, followed by mass spectrometry and spectroscopic analyses, allows verifying the effect of pitch composition and the effectiveness of solvent extraction on the separation of light and heavy components as well as the possible entrapment of light components in the heaviest insoluble fractions. The protocol set up for pitch characterization, particularly the spectroscopic analysis, is also able to individuate traces of foreign carbon particles which could interfere in the processing technologies of advanced carbon materials. In particular UV−vis, fluorescence, and infrared analyses are useful to infer properties like the aromatic and/or graphitic character of pitch components which are important in the field of pitch-based carbon fibers because affect the conductivity properties of the final product.
4. SUMMARY AND CONCLUSIONS Large, structurally different PAHs featuring coal tar pitch (CP) and naphthalene pitch (NP) fractions, separated by successive extraction with heptane and toluene, were analyzed by applying an array of many analytical and spctroscopic techniques. The thorough characterization of these pitch fractions demonstrated that the solvent extraction was key to elucidate specific structural features and relative spectroscopic signatures of smaller and larger PAH systems separately, as they otherwise overlap in the parent pitch. Overall, a crosscheck of the information obtained using different diagnostics, and in particular spectroscopic tools, was useful to distinguish among large, structurally different PAH systems featuring both pitches. From this point of view, the pitch components provide an important case study for obtaining the chemical and optical properties of large PAHs. Gas chromatographic and thermogravimetric analyses of the soluble fractions showed that the volatility of CP was mainly related to the vaporization of the PAHs that were largely present in heptane-soluble fraction and in lower amounts in the toluene-soluble fraction. The solubility of CP was mainly ruled by the molecular weight of light (from 200 up to 300 Da) and heavy (>300 Da) PAHs concentrated in the heptane and toluene fractions, respectively. Little amounts of light PAHs, which are mostly soluble in heptane, were also found in the toluene-soluble fraction possibly entrapped in the more complex matrix of higher molecular weight species (>300 Da) making up the toluene-soluble fraction. The CP coking yield could be ascribed mainly to the toluene-insoluble fraction composed of a mixture of carbonaceous species and large PAHs, with a much bigger aromatic network than the soluble fractions.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]; Tel. no.: +39-0817682254. Author Contributions
The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors accomplish this work with funding of the MSECNR projects “Miglioramento dell’efficienza energetica dei sistemi di conversione locale di energia”. I
DOI: 10.1021/acs.energyfuels.5b02576 Energy Fuels XXXX, XXX, XXX−XXX
Article
Energy & Fuels
■
(37) Ledingham, K. W. D.; Singhal, R. P. Int. J. Mass Spectrom. Ion Processes 1997, 163, 149−168. (38) Cotter, R. J. Time-of-Flight Mass Spectrometry: Instrumentation and Applications in Biological Research; American Chemical Society: Washington (USA), 1997. (39) Guillen, M. D.; Iglesias, M. J.; Dominguez, A.; Blanco, C. G. Fuel 1995, 74, 1595−1598. (40) Petrova, B.; Budinova, T.; Petrov, N.; Yardim, M. F.; Ekinci, E.; Razvigorova, M. Carbon 2005, 43, 261−267. (41) Centrone, A.; Brambilla, L.; Renouard, T.; Gherghel, L.; Mathis, C.; Mullen, K.; et al. Carbon 2005, 43, 1593−1609. (42) Rodil, S. E. Diamond Relat. Mater. 2005, 14, 1262−1269. (43) Silverstein, M., Webster, F. X, Kiemle, D. Spectrometric Identification of Organic Compounds, 4th ed.; Wiley, 2008. (44) Cristadoro, A.; Kulkarni, S. U.; Burgess, W. A.; Cervo, E. G.; Räder, H. J.; Müllen, K.; Bruce, D. A.; Thies, M. C. Carbon 2009, 47 (10), 2358−2370. (45) Clar, E. Polycyclic Hydrocarbons, Vol. 2.; Springer, 1964. (46) Rieger, R.; Mullen, K. J. Phys. Org. Chem. 2010, 23, 315−325. (47) Apicella, B.; Alfè, M.; Barbella, R.; Tregrossi, A.; Ciajolo, A. Carbon 2004, 42, 1583−1589. (48) Teo, K. B. K.; Ferrari, A. C.; Fanchini, G.; Rodil, S. E.; Yuan, J.; Tsai, J. T. H.; Laurenti, E.; Tagliaferro, A.; Robertson, J.; Milne, W. I. J. Diamond Relat. Mater. 2002, 11, 1086−1090. (49) Apicella, B.; Ciajolo, A.; Tregrossi, A. Anal. Chem. 2004, 76, 2138−2143. (50) John, P.; Soutar, I. Anal. Chem. 1976, 48 (3), 520−524. (51) Lloyd, J. B. F. Nature, Phys. Sci. 1971, 231, 64−65.
REFERENCES
(1) Alcañiz-Monge, J.; Cazorla-Amoros, D.; Linares-Solano, A. Fuel 2001, 80, 41−48. (2) Gargiulo, V.; Apicella, B.; Alfè, M.; Russo, C.; Stanzione, F.; Tregrossi, A.; Amoresano, A.; Millan, M.; Ciajolo, A. Energy Fuels 2015, 29, 5714−5722. (3) Karaca, F.; Morgan, T. J.; George, A.; Bull, I. D.; Herod, A. A.; Millan, M.; et al. Rapid Commun. Mass Spectrom. 2009, 23, 2087− 2098. (4) Herod, A. A.; George, A.; Islas, C. A.; Suelves, I.; Kandiyoti, R. Energy Fuels 2003, 17, 862−873. (5) Morgan, T. J.; George, A.; Alvarez, P.; Herod, A. A.; Millan, M.; Kandiyoti, R. Energy Fuels 2009, 23, 6003−6014. (6) Li, W.; Morgan, T. J.; Herod, A. A.; Kandiyoti R. J. Chromatogr. A 2004, 1024, 227−243. (7) George, A.; Morgan, T. J.; Alvarez, P.; Millan, M.; Herod, A. A.; Kandiyoti, R. Fuel 2010, 89, 2953−2970. (8) Esguerra, D. F.; Hoffman, W. P.; Thies, M. C. J. Supercrit. Fluids 2013, 79, 170−176. (9) Wagner, M. H.; Jäger, H.; Letizia, I.; Wilhelmi, G. Fuel 1988, 67, 792−797. (10) Diefendorf, R.; Riggs, D. 1980, U.S. Patent Number 4,208,267. (11) Ozel, M.; Bartle, K. Turk J. Chem. 2002, 26 (3), 417−424. (12) Chwastiak, S.; Lewis, I. C. Carbon 1978, 16 (2), 156−157. (13) Lewis, I. C.; Lewis, R. T. Carbon 1988, 26 (5), 757−758. (14) Zander, M. Am. Chem. Soc. Div Fuel Chem. Preprints 1989, 34, 1218−1222. (15) Greinke, R. A. Carbon 1986, 24 (6), 677−686. (16) Mochida, I.; Korai, Y.; Ku, C. H.; Watanabe, F.; Sakai, Y. Carbon 2000, 38, 305−328. (17) Morgan, T. J.; George, A.; Alvarez-Rodriguez, P.; Millan, M.; Herod, A. A.; Kandiyoti R. J. Chromatogr. A 2010, 1217, 3804−3818. (18) Morgan, T. J.; Alvarez-Rodriguez, P.; George, A.; Herod, A. A.; Kandiyoti, R. Energy Fuels 2010, 24, 3977−3989. (19) Guillen, M. D.; Blanco, J.; Canga, J. S.; Blanco, C. G. Energy Fuels 1991, 5, 188−192. (20) Karaca, F.; Islas, C. A.; Millan, M.; Behrouzi, M.; Morgan, T. J.; Herod, A. A.; et al. Energy Fuels 2004, 18, 778−788. (21) Apicella, B.; Barbella, R.; Ciajolo, A.; Tregrossi, A. Chemosphere 2003, 51, 1063−1069. (22) Alfè, M.; Apicella, B.; Barbella, R.; Tregrossi, A.; Ciajolo, A. Energy Fuels 2007, 21 (1), 136−140. (23) Rizzi, A.; Cosmina, P.; Flego, C.; Montanari, L.; Seraglia, R.; Traldi, P. J. J. Mass Spectrom. 2006, 41, 1232−1241. (24) Apicella, B.; Carpentieri, A.; Alfè, M.; Barbella, R.; Tregrossi, A.; Pucci, P.; et al. Proc. Combust. Inst. 2007, 31, 547−553. (25) Przybilla, L.; Brand, J.-D.; Yoshimura, K.; Rader, H. J.; Mullen, K. Anal. Chem. 2000, 72, 4591−4597. (26) Alfè, M.; Apicella, B.; Tregrossi, A.; Ciajolo, A. Carbon 2008, 46, 2059−2066. (27) Russo, C.; Stanzione, F.; Tregrossi, A.; Ciajolo, A. Carbon 2014, 74, 127−138. (28) Morris, J. V.; Mahaney, M. A.; Huber, J. R. J. Phys. Chem. 1976, 80 (9), 969−974. (29) Berlman, I. B. Handbook of Fluorescence Spectra of Aromatic Molecules, 2nd ed.; Academic Press: New York, 1971. (30) Fetzer, J. C. Large (C> = 24) Polycyclic Aromatic Hydrocarbons: Chemistry and Analysis; Wiley, 2000. (31) Morgan, T. J.; George, A.; Alvarez, P.; Millan, M.; Herod, A. A.; Kandiyoti, R. Energy Fuels 2008, 22 (5), 3275−3292. (32) Guillen, M. D.; Iglesias, M. J.; Dominguez, A.; Blanco, C. G. Energy Fuels 1992, 6 (4), 518−525. (33) Guillén, M. D.; Domínguez, A.; Iglesias, M. J.; Blanco, C. G. Fuel 1995, 74 (2), 233−240. (34) Guillén, M. D.; Domínguez, A.; Iglesias, M. J.; Fuente, E.; Blanco, C. G. Fuel 1996, 75 (9), 1101−1107. (35) Lewis, I. C. Fuel 1987, 66, 1527−1531. (36) Herod, A. A.; Bartle, K. D.; Morgan, T. J.; Kandiyoti, R. Chem. Rev. 2012, 112, 3892−3923. J
DOI: 10.1021/acs.energyfuels.5b02576 Energy Fuels XXXX, XXX, XXX−XXX
