Indian Journal of Chemistry Vol. 40A, April 2001, pp. 374-379
Characterization and investigation of polycyclic aromatic compounds present in petrol, diesel, kerosene and 2T oil using excitation emission matrix fluorescence
Digambara Patra, Lakshmi Sireesha K & A K Mishra*
Department of Chemistry, Indian Institute of Technology, Chennai 600 036, India Received 5 April 2000; revised 7 August 2000
Excitation emission matrix fluorescence fingerprint has been used for characterization of motor oils like diesel, petrol, kerosene and lubricant oil (2T oil). Heavy oil like diesel and 2T oil contain mostly higher membered polycyclic aromatic compounds, whereas lighter oils like petrol and kerosene contain intermediate polycyclic aromatic compounds along with their lower membered aromatics. Analysis of synthetic mixture of diesel and kerosene has been carried out to check contamination of diesel qualitatively by kerosene. A multivariate method to estimate kerosene and diesel in their mixture comprising their dilute solutions has been used, which gives a satisfactory result. Estimation of 2T oil (0-10%, v/v) in petrol has been successfully carried out using this technique.
Petroleum products like diesel, kerosene, petrol, 2T oil, etc., contain many polycyclic aromatic compounds (PACs), which differ in their types and kind according to the application of petroleum products'. These PACs are highly fluorescent and, therefore, fluorescence measurements, in contrast to conventional methods, can be performed directly to differentiate between them. However, higher concentration of PACs and their multiplicity in petroleum products leads to energy transfer among PACs and makes the analysis difficult2-4
. Therefore, conventional fluorescence measurement never works in the case of motor oils2 and multi-wavelength excitations5 are needed to characterize these oils.
In the last few years, development in fluorescence instrumentation has led to new approaches in tluorimetric determination of multi-component systems. Among them, Excitation Emission Matrix Fluorescence (EEMF)6•7, which is also known as total fluorescence spectroscop/, is a rapid and inexpensive technique used for the analysis of polycyclic aromatic hydrocarbons9'10 and petroleum products8·''·'2 in the environment. This technique allows plotting of emission intensities at all combinations of excitation and emtsston wavelengths in a single three- dimensional array either as a contour diagram or as a topographical surface6·7• The EEMF analysis provides a "fingerprint" consisting of a 3-D emission/excitation intensity contour diagram. This "fingerprint" is used for qualitative information about the content and
*Corresponding author. Tel (044) 4458251. Fax: (044) 2352545 E-mail: mishra @acer.iitm.ernet.in
relative composition of PACs in the sample.
In this work, using EEMF for characterization of motor oils, like diesel, petrol, kerosene and 2T oil, the compositional variability and semi-quantitative analysis of these oils have been investigated. Since contamination of diesel by kerosene is a serious problem in South Asia13-19 and fluorescence provides high sensitivity and selectivity along with simplicity, EEMF has been tested as an analytical technique to check contamination of diesel qualitatively in the synthetic mixtures of kerosene and diesel. A multivariate method20 has also been applied to determine quantitatively amounts of kerosene and diesel present in their mixture composed of dilute solutions of these components, where it is expected that there is no energy transfer among the PACs. The method holds good and gives a good recovery. The quantitative analysis of 2T oil present in petrol has also been carried out using this technique.
Materials and Methods
Samples of diesel, kerosene, petrol and 2T oil (Servo) were collected from the local market in Chennai and their purity was tested before analysis.
Cyclohexane (HPLC grade, Ranbaxy) was used as a solvent after purification and was subjected to blank experiments to ensure its fluorimetric purity.
Measuring known volumes of the desired motor oils (neat samples of kerosene and diesel) and mixing them in different proportions, the synthetic mixture of motor oils were prepared. To check the spectroscopic validity, test samples from ten different places in Chennai were collected within a time gap of 6
months. Except a minor change in fluorescence intensity, no variation in other fluorescence properties was observed for all these samples.
Experimental technique
Fluorescence spectra were obtained on a Hitachi F- 4500 spectrofluorimeter. For EEMF measurement, the scan speed was 240 nm/sec and PMT voltage was at 700 V. Excitation and emission slit widths were 5 nm. Excitation source was 100 W Xenon lamp. The EEMF spectra was recorded in the excitation wavelength range 250-500 nm and emtsston wavelength range 300-600 nm with an interval of 5nm each for diesel, kerosene, 2T oil and petrol. A right angle geometry was used for measurement. For neat samples where the optical densities are high, the spectra reflect a combined effect of fluorescence and inner filter effect.
Results and Discussion
Qualitative study on petroleum products
Generally, polycyclic aromatic compounds emit in
Fig I a / EEMF or Kerosene (Neat)
Excitation wavelength (run)
600~---, Fig lb EEMF of Kerosene (neat)
550
~ i
500i
450~
.
8"'
· i
400I'll 350
300
250 300 350 500
Excitation wavelength (nm)
Fig. 1 - EEM fluorescence of neat sample of kerosene. [(a) topographical surface diagram, (b) contour map. The arrow mark indicates the response from the class of PACs (aromatic ring size) present in the sample].
the higher energy range, and with an increase in the number of benzene rings the emission maxima of aromatic compounds are shifted towards longer wavelengths. Two-ring aromatic hydrocarbons such as naphthalene exhibit fluorescence maxima at low emission and excitation wavelengths, while five-ring compounds such as perylene have maxima at high emission and excitation wavelengths. Three- and four- ring compounds exhibit intermediate fluorescence characteristics21·22. A substitution of the aromatic compounds by alkyl, phenyl, or other functional groups has nearly no influence on the position of the absorption spectra21·22. Gray et al.8 have studied the EEMF fingerprint region of different types of crude oils with respect to their polycyclic aromatic hydrocarbons content. The EEMF fingerprint of neat sample of kerosene is given in Fig. 1. It shows that kerosene mostly contains intermediate PACs like three-ring systems along with smaller ring systems.
The EEMF fingerprint of petrol (Fig. 2) also shows that petrol contains mostly three-ring aromatics along with the smaller ring systems.
EEMF of Petrol (Neat) Fig 2a
Excitation wavelength (nm)
600
550
1
.c 500 lj,.§
!.! 450
;
"Ill 8 400
i
I'll 350300 250
Flg2b
300
EEMF of Petrol (neat)
350 400
&citation waveJen&th (nm)
450 soo
Fig. 2-EEM fluorescence of neat sample of petrol. [(a) topographical surface diagram, (b) contour map. The arrow mark indicates the response from the class of PACs (aromatic ring size) present in the sample].
376 INDIAN J. CHEM. SEC. A, APRIL 2001
But a neat sample of diesel shows two different bands in the 3D-emission/excitation contour map (Fig. 3). This 3D-emission/excitation contour map gives the distribution pattern of various classes of aromatic compounds present in these systems. It is clear form Fig. 3 that band I is due to six-ring aromatics, whereas band II is due to five-ring aromatics. This observation is in accordance with the previous quantitative analysis carried out for the diesel sample, which tells that diesel has an average composition of 16.5 wt% mono-, 1 1.6 wt% di-, 0.5 wt% tri-, and <<0.1 wt % polycyclic aromatic hydrocarbons23. Recently, using Time-of-Flight Mass Analysis, Hankin and John have also found that, in general, diesel contains six-membered ring aromatic compounds along with five-, four-, three-, two- and
b d . . d ?4
one-mem ere nng aromatiC compoun s- .
I band-....__ EEMF <1 Diesel (neat) Fig3a --...
650
Excitation wavelength (nm) 650
Fig3b 600
!
550.c
i
5004i
.. ..
~ EEMF <I Diesel (neat)=
450"Dl 0
-~ 400 w
350
300 350 400 450 500 550
Excitation wavelength (om)
Fig. 3-EEM nuorescence of neat sample of diesel. [(a) topographical surface diagram, (b) contour map. The arrow mark indicates the response from the class of PACs (aromatic ring size) present in the sample].
2T oil gives an EEMF fingerprint of four- and five- ring systems (Fig. 4). Therefore, EEMF fingerprint distribution indicates that heavy oil, like diesel and 2T oil, contains mostly higher membered ring PACs and lighter oil, like kerosene and petrol, contains intermediate PACs along with their smaller membered ring PACs. The absence of fluorescence from lower membered PACs present in petroleum products is partly due to self-quenching of fluorophores at higher concentration and partly because of resonance energy transfer from lower aromatics to higher aromatics2·4
. Resonance energy transfer between molecules occurs by a Foster type mechanism when the emission spectrum of the energy donor overlaps with the acceptor absorption spectrum25.
Fig4a
~of2Toil(neat)
Excitation wavelength (om)
600
Flg4b 550
!
EEMF or2T oil (neat) .c 500j .. .. ..
450 II:j
400350 300
250 300 350 400 450 500
E:ldtatloo wavelength (om)
Fig. 4-EEM nuorescencc of neat sample of 2T oil. f(a) topographical surface diagram, (b) contour map. The arrow mark indicates the response from the class of PACs (aromatic ring size) present in the sample].
Table !-Composition of synthetic mixture of kerosene and diesel at dilute conditions and their recovery by EEMF method Sl. No. Cone. of diesel Cone. of diesel % Recovery of
taken (g/ml) found (g/ml) diesel I I X 10'5 1.09 X 10·5 ± 0.0 I 108.8
2 8x 10·6 8.26 X 10·6 ± 0.01 103.3
3 6x 10·6 6.61 X 10'6 ± 0.01 110.0
4 2 X 10'5 1.90x 10·5 ± 0.01 95.2
5 6x 10·5 4.79 X 10'5 ± 0.01 79.9
Qualitative analysis of synthetic mixture of diesel and kerosene
The synthetic mixture of diesel and kerosene has been studied qualitatively using EEMF. Fig. 5 shows the EEMF fingerprint of a synthetic mixture of kerosene and diesel. It is clear from the figure that as contamination by kerosene proceeds the fluorescence intensity at band I (six-ring aromatics) decreases thereby increasing the fluorescence intensity at band II (five-ring aromatics), which is, in fact, due to resonance energy transfer. This is well reflected in Fig. Sa for I 0% of contamination, in which the relative fluorescence intensity at band I (six-ring aromatics) decreases and that of band II increases (five-ring aromatics) compared to that of neat diesel sample. Fig. Sb shows EEMF for 50% of contamination where the relative fluorescence intensity at band II (five-ring aromatics) increases by decreasing the fluorescence of band I, and a new band (III band) appears at lower wavelength which reflects the emission for four-ring systems. Fig. Sc shows EEMF for 90% contamination, which shows the fluorescence from only band Ill (four-ring aromatics) suppressing all the fluorescence from bands I & II.
Synthetic mixtures of diesel and petrol also show similar behaviour. The insenstttve nature of fluorescence properties of kerosene in contaminated mixture ts because of the self-quenching of fluorophores present in kerosene at higher concentration and the variation of fluorescence properties of diesel with dilution (contamination) is due to the resonance energy transfer from lower to higher aromatics2-4
. Therefore, by monitoring the change in fluorescence intensity in the EEMF fingerprint at bands I, II & III contamination of diesel can be checked qualitatively.
Quantitative analysis of diesel and kerosene
Additivity of fluorescence intensity at higher concentration cannot be tested since the PACs present in kerosene and diesel undergo energy transfer processes. Therefore, quantification has been carried
Cone. of kerosene Cone. of kerosene found % Recovery of
taken (g/ml) (g/ml) kerosene
I X 104 0.74 X I 04 ± 0.02 71.4 I X 10'5 0.71 X 10-5± 0.02 71.0
4x 104 3.62 X 104 ± 0.08 90.4
2x 10·5 2.02 X I 0·5 ± 0.09 101.0 I X 10'5 1.05 X 10.5+0.09 105.3
EEMF or kerosene In diesel (10 %, v/v) I
Fig Sa
Exdtatloo wavelength (om)
EEMF or kerosene In diesel (SO %, v/v)
Fig Sb I band
m
Exdtatioo wavelength (om)
EEMF or kerosene In diesel (90 %, v/v) IUbaod
Exdtatioo wavelength (om)
Fig. 5-EEM Ouorescence of synthetic mixture of kerosene in diesel f(a) 10% (v/v), (b) 50% (v/v) and (c) 90% (v/v). The arrow mark indicates the response from the class of PACs (aromatic ring size) present in the sample].
378 INDIAN J. CHEM. SEC. A, APRIL 2001
out in dilute solutions. Before analysis, it was ensured that there was no energy transfer, there was no charge transfer and no quenching occurred between different fluorophores. It was observed that this condition would hold only when the analyte concentration was small, i.e., in dilute solution. Therefore, the observed fluorescence intensity of the mixture solution was the sum of the individual components present in the mixture.
Thus, for a two-component system, fluorescence intensity of the mixture at various excitation/emission wavelengths Mj obtained from EEMF can be written as2o
'
... (1)
i=l
where 'i' is the number of components present in the mixture; the coefficient Ci is a ratio of the concentration of the ith component in the mixture to the concentration of its standard stock; Mij is the fluorescence intensity of the ith component at various possible excitation/emission wavelengths U) of standard stock solution. Since the number of data points in the above equation is greater than number of components present, therefore, to solve this equation, the matrix can be written as
... (2)
Number of data points (j) for each EEMF spectral region used in the calculations after filtering the scattering light and unwanted spectra was at least 1000. To solve this, a C-Janguage based program was written in our Jab. The standard solution for kerosene
Table 2-Statistical parameters calculated for the estimation of kerosene and diesel in their synthetic mixtures at dilute
conditions and for 2T oil present in petrol Statistical Diesel and kerosene 2T oil in petrol parameter mixture (g/ml) (in%, v/v)
Diesel Kerosene ByYM By EEMF SEP 6.09x 10-6 2.31 x
w -
5 0.197 0.100RMSD 5.45 x
w-
6 2.06x 10·5 0.170 0.087REP(%) 26.2 19.1 4.1 2.0
and diesel was chosen as 3xl0·5 g/ml and 3xl0·6 g/ml respectively. The limit of quantification for kerosene and diesel by the method is found to be 1 X I 0-5 g/ml and 5xl0·6 g/ml respectively. The results obtained and the recovery found out are given in Table I. The results are quite satisfactory.
For the above estimation, statistical parameters like standard error of prediction (SEP), root mean square deviation (RMSD) and relative eiTor of prediction (REP (%)) were obtained. The equations for the statistical parameters SEP, RMSD and REP(%) are:
SEP = [L, (x-xc)2/(n-l)] 112 RMSD = [L, (x-xc) 2 In] 112
REP (%)=(100/xa)
[L
(x-xc)2/ n]112where x is the real concentration and Xc the estimated concentration of one of the analyte in a series of samples, Xa is the average real concentration of the analyte in a series of sample and n corresponds to the number of samples evaluated. The calculated statistical parameters of these mixture samples are given in Table 2.
Estimation of2T oil in petrol
The estimation of 2T oil in petrol was done in the 2T oil concentration range 1 to 10%, v/v by EEMF.
The excitation and emission wavelength where neat sample of 2T oil gives the maximum fluorescence intensity
0.-ex
= 465 nm &A em
= 500 nm) was chosen for fluorescence intensity measurements for calibration. The linear regression (r) found out from the calibration plot was 0.997 (slope= 24.85, intercept= 19.58). The concentrations in unknown samples were found out from the calibration plot of 2T oil in petrol and compared with that of viscosity measurement (VM) (r=0.99, slope=23.35xl0·2, intercept=0.71544). The results obtained by the method are good and recovery was found to be more than 95%. The results are summarized in Table 3 and the calculated statistical parameters are given in Table 2.
Table 3-Comparison of quantification of 2T oil present in petrol by EEMF and viscosity measurement Sl. No. Cone. of Cone. of %Recovery Cone. of 2T oil % Recovery by
2T oil taken 2T oil found by VM by YM found by EEMF EEMF
(in%, v/v) (in%, v/v) (in %, v/v)
2.5 2.8±0.2 116.6 2.6±0.2 104
2 3.5 3.6±0.1 105.8 3.44±0.1 98.3
3 4.5 4.42±0.06 100 4.42± 0.06 98.2
4 6.0 5.9±0.1 98.3 5.9±0.1 98.3
Concluding remarks
The use of EEMF in the analysis of petroleum fuels and 2T oil is seen to offer an easy and inexpensive method. Both multivariate as well as calibration method are found to be satisfactory in the estimation of components in mixtures.
References
I Gruse W A & Stevens D R, Chemical technology of petroleum (McGraw Hill, New York), 3'd Ed., 1960,425. 2 Patra D, Sireesha K L & Mishra A K, J scient ind Res, 59
(2000) 300.
3 Wang X & Mullins 0 C, Appl Specrtosc, 48 (1994) 977.
4 Ralston C Y, Yu X & Mullins 0 C, Appl Spectrosc, 50 (1996) 1563.
5 Bublitz J & Schade W, SPIE, 2504 ( 1996) 265.
6 Rho 1 H & Stuart J L, Anal Chem, 50 ( 1978) 620.
7 Senga Y & Minamin S, Appl Spectrosc, 45 ( 1991) 1721. 8 Gray N R, McMillen S J, Requejo A G, Kerr J M, Denoux G
& McDonald T J, SPE, 25990 (1993) 465.
9 Beltran J L, Ferrer R & Guiteras J, Anal Chim Acta, 373 ( 1998) 311.
10 JiJi R D, Cooper G A & Booksh K S, Anal Chim Acta, 397 ( 1999) 61.
II Dudelzak A E, Babichenko S M, Poryvkina L V & Saar K J, Applied Optics, 30 (1991) 453.
12 Shanahan G J, BoleyN P & Traynor AD, Appl Spectrosc, 45 (1991) 1070.
13 Gupta A A, Swami K K, Misra A K, Bhatnagar A K &
Mukhopadhaya P, Hydrocarbon Technology, 15 (1992) 137.
14 Bhatnagar V P, J Acoust Soc India, 9 ( 1981) 19.
15 Shahru Bahari M, Criddle W J & Thomas J DR, Analyst, 115 (1990) 417.
16 Shahru Bahari M, Criddle W J & Thomas J D R, Anal Proc, 28 (1991) 14.
17 Shahru Bahari M, Criddle W J & Thomas J D R, Anal Proc.
29 ( 1992) 30.
18 Srivastava A, Shukla R K, Srivastava A & Bali L M, Optical sensor for determining adulteration in petrol by kerosene, Proc. Inti. Conf On Fiber Optics and Photonics, Vol. II, (Tata McGraw Hill, New Delhi), 1997,989.
19 Roy S, Sensors & Actuators B, 55 (1999) 212.
20 Patra D & Mishra A K, Anal Lett, 33 (2000) 2293.
21 Berlman I B, Handbook of fluorescence spectra of aromatic molecules (Academic Press, New York), Chap. I, 1965, 14.
22 Birks 1 B, Photophysics of aromatics molecules (Wiley- lnterscience, London), 1972.
23 Wenck H & Schnieder Ch., DGMK:Project 409, Collection of chemical and physical data of automotive fuels. 12-19.
(Deutsche Wissenschaftliche Gesellschaft fUr Erdol, Erdgas und Kohle e.v., Hamburg), 1993.
24 Hankin S & John P, Anal Chem, 71 ( 1999) II 00.
25 Lakowicz J R, Principles of fluorescence spectroscopy (Plenum Press, New York), 3'd Ed., Chap. 7, 1986, 190.
