Identification of unknown compounds by using a novel retention index system in liquid chromatography

Abstract

A method for the identification of unknown compounds based on a novel Retention Index System having a TAGs homologous series, wherein such identification is performed by means of liquid chromatography (LC), or liquid chromatography coupled with mass spectrometry (LC-MS) is disclosed.

Claims

1. A method for the identification of unknown compounds in liquid chromatography based on a homologous series of triacylglycerols (TAGs) composed by odd-chain fatty acids as a Retention Index System comprising the following steps: a. attribution of a retention index value to each reference TAG; b. chromatographic separation of a TAGs reference mixture; c. chromatographic separation of unknown compounds in a sample; d. determination of the retention index value of each separated unknown compound; and e. identification of each unknown compounds.

2. The method of claim 1, wherein the homologous series comprises triacylglycerols ranging from triheptanoin to triheneicosanoin, or any portion thereof.

3. The method of claim 2, wherein the homologous series comprises triacylglycerols ranging: from triheptanoin (C7C7C7) to trinonadecanoin (C19C19C19); or from trinonanoin (C9C9C9) to trinonadecanoin (C19C19C19); or from trinonanoin (C9C9C9) to triheneicosanoin (C21C21C21); or any portion thereof.

4. The method of claim 1, wherein the chromatographic separation (steps b. and c.) is performed in an acetonitrile/2-propanol mobile phase.

5. The method of claim 4, wherein the chromatographic separation is performed in gradient elution.

6. The method of claim 1, wherein the determination of the retention index value (step d.) is carried out by applying the following equation: $\begin{array}{cc}LRl=100\left[z+n\frac{t_{Ri}-t_{Rz}}{t_{R\left(z+n\right)}-t_{Rz}}\right]& \end{array}$ wherein, t.sub.Ri is the retention time of the unknown compound i, t.sub.Rz and t.sub.R(z+1) are the retention times of the TAGs which elute, respectively, immediately before and immediately after the unknown compound i, z is the partition number (PN) associated with the compound eluted immediately before the unknown compound i, and n represents the difference in z units between the reference TAGs eluted immediately before and after the unknown compound.

7. The method of claim 6, wherein the identification step e. is performed by comparison of the calculated LRI value for the unknown compound with those previously obtained for listed compounds.

8. The method of claim 6, wherein the identification step e. is automatically performed by means of a software, wherein said software automatically matches the LRI value of the unknown compound with those of a LRI library.

9. The method according to claim 1, wherein the obtained retention index values are used to discriminate between analytes with the same mass fragmentations.

10. The method according to claim 1, wherein the obtained retention index values have a determination repeatability minor than the peak width at base of the unknown compounds.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1: UHPLC-ELSD method development steps: A) chromatogram of cod liver oil under isocratic conditions (40% solvent B); B) chromatogram of cod liver oil under 70 min linear gradient (20-50% B) with 5 min isocratic step (20% solvent B); C) chromatogram of cod liver oil under multi-step gradient elution (0-25% B in 30 min and 25-50% B in 75 min); D) chromatogram of butter sample under multi-step gradient elution (0-25% B in 30 min and 25-50% B in 75 min); E) chromatogram of butter sample under 105 min linear gradient (0-50% B).

(2) FIG. 2: Calibration curve t.sub.R vs PN for the reference TAGs comprised in one of the homologue series (from C9C9C9 to C19C19C19).

(3) FIG. 3: UHPLC-ELSD selected chromatograms acquired by using the final developed method: A) borage oil; B) goat milk; C) menhaden oil; D) reference homologue series.

(4) FIG. 4: Chromatograms of identified A) linseed oil and B) menhaden oil.

(5) FIG. 5: ESI-MS spectrum of A) LLP and B) OOLn, with fragment elucidation. MAGs: monoacylglycerols; DAGs: diacylglycerols; TAG+Na+: sodium adduct of triacylglycerol.

(6) FIG. 6: Chromatogram of the selected homologous series, together with both SN and ΔLRI values (Δ) for each pair of adjacent homologues.

(7) FIG. 7: Histogram reporting the average LRI values between UHPLC-ELSD and HPLC-ESI-MS analyses (values are reported in Table 8); error bar corresponds to ΔLRI value.

(8) FIG. 8: sea bass chromatogram, provided as example of a sample chromatogram identified using solely the LRI database.

DEFINITIONS

(9) The term “homologous series” refers to a series of compounds with the same general formula varying only in the number of methylene groups in an alkyl chain. In other words, successive members differ in mass by an extra methylene bridge (i.e. —CH.sub.2—) inserted in the chain. Adjacent members in this series are known as “adjacent homologues”.

(10) For the purpose of the present invention, when triacylglycerols composed by odd-chain fatty acids are preferred, adjacent members (or adjacent TAGs) such as, for example, trinonanoin (C9C9C9) and triundecanoin (C11C11C11), differ in mass by six extra methylene bridges.

(11) The term “reference mixture” refers to the group of compounds that are part of the homologous series selected as standards for the purpose of the present invention.

DETAILED DESCRIPTION OF THE INVENTION

(12) A first object of the present invention is a homologous series of triacylglycerols for use in liquid chromatography, wherein said series preferably comprises simple triacylglycerols composed by odd-chain fatty acids.

(13) Triacylglycerols were selected as main target analytes because of their biological importance and very regular liquid chromatographic profile, especially in reverse phase liquid chromatography (RP-LC), and since they have never been considered before in similar studies.

(14) Here below a triacylglycerol general structure is reported:

(15) ##STR00001##

(16) wherein,

(17) R′, R″ and R′″, taken together or alone, can be the same or a different straight alkyl chain.

(18) Triacylglycerols of the present invention can be simple or mixed TAGs. Simple TAGs comprise only one kind of fatty acid (i.e. R′, R″, R′″ are the same) while mixed TAGs comprise two or three different fatty acids (i.e. two or three among R′, R″, R′″ are different). Preferably, reference triacylglycerols of the homologous series are simple TAGs. Each fatty acid can be saturated or unsaturated and may contain one, two or three unsaturations. Preferably, reference triacylglycerols of the homologous series are saturated TAGs. Triacylglycerols of the present invention may comprise both even- and odd-chain fatty acids. Preferably, reference triacylglycerols of the homologous series comprises odd-chain fatty acids.

(19) Reference triacylglycerols of the present invention are commercially available. Alternatively, they can be readily synthetized either by esterification reaction of glycerol with three suitable fatty acids or by transesterification reaction of glycerol with three suitable esters.

(20) Since the homologous series employed has to cover the expected retention times of all possible analytes, reference triacylglycerols of the present invention have been selected so that they cover a wide range of retention times.

(21) In one embodiment of the present invention, the reference homologous series comprises, but is not limited to, triacylglycerols ranging from triheptanoin (C7C7C7) to triheneicosanoin (C21C21C21) or any portion thereof.

(22) For the purpose of the present invention in fact, any portion of said homologous series of adjacent TAGs is suitable as reference mixture, provided that it comprises at least two adjacent TAGs selected from the group consisting of, but not limited to, the followings: triheptanoin (C7C7C7), trinonanoin (C9C9C9), triundecanoin (C11C11C11), tritridecanoin (C13C13C13), tripentadecanoin (C15C15C15), triheptadecanoin (C17C17C17), trinonadecanoin (C19C19C19) and triheneicosanoin (C21C21C21).

(23) In a preferred embodiment of the present invention, the reference homologous series comprises triacylglycerols ranging: from triheptanoin (C7C7C7) to trinonadecanoin (C19C19C19); or from trinonanoin (C9C9C9) to trinonadecanoin (C19C19C19); or from trinonanoin (C9C9C9) to triheneicosanoin (C21C21C21); or any portion thereof.

(24) In particular, the reference mixture ranging from C9C9C9 to C19C19C19 largely covers the chromatographic space in which lipid samples, relevant from an alimentary point of view, generally fall. More in particular, they cover almost all the chromatogram, from about 9 to about 120 min with a homogeneous distribution, as shown in FIG. 3D.

(25) In addition, the reference mixtures above disclosed can be extended either to shorter-chain triacylglycerols (e.g., to triformin, C1C1C1) or longer-chain triacylglycerols according to the chromatographic method and/or sample composition employed.

(26) Mainly due to its optimal coverage of chromatographic space and to the fact that it is not affected by chromatographic analytical conditions, the reference TAGs homologous series makes the Retention Index System of the present invention reproducible both at intra- and inter-laboratory levels.

(27) Further provided herein is a kit for use in liquid chromatography comprising a homologous series of triacylglycerols.

(28) Preferably, the kit comprises a reference homologous series comprising simple triacylglycerols composed by odd-chain fatty acids.

(29) In one embodiment, the kit comprises a reference homologous series comprising, but not limited to, triacylglycerols ranging from triheptanoin (C7C7C7) to triheneicosanoin (C21C21C21), or any portion thereof.

(30) For the purpose of the present invention, the kit can comprise any portion of said homologous series, provided that it comprises at least two adjacent TAGs selected from the group consisting of, but not limited to, the followings: triheptanoin (C7C7C7), trinonanoin (C9C9C9), triundecanoin (C11C11C11), tritridecanoin (C13C13C13), tripentadecanoin (C15C15C15), triheptadecanoin (C17C17C17), trinonadecanoin (C19C19C19) and triheneicosanoin (C21C21C21).

(31) In a preferred embodiment of the present invention, the kit comprises a reference homologous series comprising, but not limited to, triacylglycerols ranging: from triheptanoin (C7C7C7) to trinonadecanoin (C19C19C19); or from trinonanoin (C9C9C9) to trinonadecanoin (C19C19C19); or from trinonanoin (C9C9C9) to triheneicosanoin (C21C21C21); or any portion thereof.

(32) In addition, the kit of the present invention can comprise a homologous series such that each of the reference homologous series above disclosed, or a portion thereof, can be extended either to shorter-chain triacylglycerols (e.g., to triformin, C1C1C1) or longer-chain triacylglycerols, according to the chromatographic method and/or sample composition employed.

(33) Said kit can comprise the TAGs homologous series either in the form of individual vials, each comprising a reference TAG, or as a unique vial comprising the TAGs reference mixture according to the present invention. Each reference TAG and/or the TAGs reference mixture can be in a solid dried form or as a ready-to-use solution to be introduced into the LC system previously or together with the unknown compound (i.e. analyte).

(34) Preferably, the kit provides the TAGs reference mixture as a ready-to-use solution in a suitable solvent, to be introduced in the LC system previously or together with the unknown compound, after a simple dilution in a suitable solvent.

(35) Practically, the kit provides the TAGs reference mixture as a ready-to-use solution in 2-propanol at a concentration of, but not limited to, 1000 mg/L for each TAG, to be introduced in the LC system previously or together with the unknown compound (i.e. analyte), after a simple dilution in 2-propanol.

(36) The standard mixture of selected triacylglycerols (i.e. homologous series) is preferably prepared in a concentration range 50-1000 mg/L, depending on the employed detector, viz. on its sensitivity: a concentration of 50-100 mg/L is used when MS detector are employed, while higher concentrations are employed for less sensitive ELSD or UV detectors. Then, the homologous series can be used both as external and internal standard.

(37) In the first case, such TAGs standard mixture is injected into a liquid chromatography system in order to separate each reference triacylglycerol. The resulting chromatogram forms the reference pattern of triacylglycerols useful as external standard for subsequent qualitative analysis.

(38) The reference triacylglycerols of the present invention can also be used as internal standard by co-injection with the sample whose analytes are to be identified. In case of very complex matrices, however, the external standards approach should be preferred in order to avoid the chromatogram to be over-populated also with reference peaks. This, in fact, may pose a problem of resolution.

(39) A further object of the present invention is the development of a method for the identification of unknown compounds in liquid chromatography based on a TAGs Retention Index System, wherein said method is in particular a high performance liquid chromatography (HPLC) or an ultra-high performance liquid chromatography (UHPLC) method able to maximize the baseline separation of lipid compounds in real-world samples.

(40) More specifically, a preferred object of the present invention is the development of an UHPLC method able to maximise the baseline separation of lipid compounds in real-world lipid samples effective and reliable for the univocal identification of unknown compounds with, and even without, the complementarity of mass spectrometry.

(41) Therefore, a method for the identification of unknown compounds in liquid chromatography is herein described, the method comprising the following steps: a. attribution of a retention index value to each reference TAG; b. chromatographic separation of a TAGs reference mixture; c. chromatographic separation of unknown compounds in a sample; d. determination of the retention index value of each separated unknown compound; e. identification of the unknown compound.

(42) Since it allows for the conversion of retention times into system-independent constants, the novel Retention Index System here disclosed can be used independently from the selected instrumentation and analytical conditions.

(43) With respect to step a., as above disclosed, it is known from the literature that a retention index value can be arbitrarily attributed to an organic compound according to the number of carbon atoms in the alkyl chain thereof.

(44) Therefore, for the purpose of the present invention, to each reference TAG of the claimed homologous series, a retention index value was arbitrarily attributed according to the following:
LRI.sub.std=100×PN

(45) wherein,

(46) PN (Partition Number) is related to the carbon chain length (CN) and double bond number (DB) of the fatty acids bound to the glycerol backbone, according to the following equation:
PN=CN−2DB

(47) By way of example, LRI.sub.std is equal to 2100 for triheptanoin (C7C7C7), 2700 for trinonanoin (C9C9C9), 3900 for tritridecanoin (C13C13C13), 5700 for trinonadecanoin (C19C19C19) etc.

(48) In detail, the chromatographic separation of both step b. and c. can be carried out by using either an HPLC or an UHPLC system. In the first case, analyses were carried out by using an HPLC system, for example, a Waters Alliance HPLC system (Waters Associates Inc) optionally coupled to a mass spectrometer, such as an electrospray ionization mass spectrometer (i.e. ESI-MS). In the second case, analyses were carried out by using an UHPLC system, for example, a Nexera X2 system (Shimadzu) coupled with a suitable detector. For the purpose of the present invention suitable detectors can be destructive or non-destructive detectors selected among, but not limited to, the following: evaporative light scattering detector (ELSD), ultraviolet detectors, both UV and photodiode array detectors, mass spectrometers by both atmospheric pressure interfaces (API) or electron ionization (EI) source. The ELSD is preferred. Additionally, in the event that the chromatograph is coupled with a mass spectrometer, suitable API interfaces can be electrospray (ESI), atmospheric pressure chemical ionization (APCI) or atmospheric pressure photoionization (APPI), all suitable for TAGs analysis.

(49) As regard the analytical conditions, the inventors of the present invention have succeeded in establishing a method almost independent from both the chromatographic instrumentation and the conditions employed.

(50) The only feature that affects the method and, in particular, the method reproducibility is the mobile phase. In fact, for the chromatographic separation of both step b. and c., the selected mobile phase comprises acetonitrile (hereinafter also referred as solvent A) and 2-propanol (hereinafter also referred as solvent B). The inventors, however, have surprisingly found that this mobile phase (i.e. acetonitrile/2-propanol mobile phase), operated in gradient elution, preferably in linear gradient elution, allows to obtain an optimal compromise between chromatographic resolution (i.e. a peak capacity value of about 162) and analysis time (i.e. about 125 min), as far as we know, the best ever reached so far for very complex samples, such as fish oil and dairy products.

(51) Furthermore, the toxicity of both acetonitrile and 2-propanol is by far lower than other solvent employed in previous methods, such as dichloromethane or acetone, characterized inter alia from very high volatility.

(52) Another issue was the selection of the maximum percentage of the strongest solvent (i.e. 2-propanol). After extensive experimentation, the inventors found that, under isocratic conditions, a percentage of 2-propanol between 40% and 60%, preferably between 45% and 55%, and more preferably equal to 50%, was sufficient to completely elute all the compounds (FIG. 1A) in one of most complex real-world samples, namely a cod liver oil. In order to improve the chromatographic method, a gradient elution program was then carefully evaluated aiming to maximise the baseline separation of complex real samples.

(53) Suitable gradient programs have been settled up, in particular the following: i. 0-5 min, 20% B, 75 min to 50% B (FIG. 1B) which however shows a poor resolution at the low partition number (PN) regions (i.e. from about 30 to about 36) and the presence of several co-elutions in the richest PN region, namely PN 42. ii. 0-30 min, 0-25% B, 30-105 min to 50% B (FIG. 10) by means of which a clear separation between 30, 32 and 34 PN regions was achieved, as well as a major resolution in each PN region; and finally iii. a linear gradient 0-50% B in 105 min (FIG. 1E).

(54) The latter, i.e. iii., resulted in a great improvement of the present chromatographic method reaching a better resolution, an improved peak shape, showing more uniform and narrower peak widths (see FIG. 1E with respect to FIG. 1D), and an optimal time of analysis.

(55) For these reasons, such linear gradient acetonitrile/2-propanol represents the best suited to the method of the present invention, in particular with respect to the chromatographic separation of both step b. and c.

(56) With the aim of maximizing the separation, the inventors also took carefully into consideration the temperature. Three temperatures were evaluated, in the range 30-40° C. (i.e. 30° C., 35° C., and 40° C.), leading to the conclusion that, in said range, the method is readily reproducible. However, a temperature of 35° C. is preferred. The experiments provided herein show also that the method of the present invention is independent from analytical conditions and chromatographic instrumentation such as, for example, column dimensions, stationary phase packing, flow rate and gradient steepness of the mobile phase (see Evidence of the method independency: repeatability studies).

(57) The method claimed is as independent as possible from analytical conditions allowing to conclude that the Retention Index System of the present invention, and the database based thereon, are usable in samples identification at both intra- and inter-laboratory level.

(58) The reference mixture of step b. preferably comprises simple triacylglycerols composed by odd-chain fatty acids which represent the reference homologous series of the novel Retention Index System according to the present invention.

(59) Step b. can further comprise the eventual set up of a calibration curve t.sub.R vs. PN. In detail, given the linear relationship between t.sub.R and PN, it is possible to build a plot using the PN value of the reference TAGs in step a. and the retention times obtained in step b (FIG. 2).

(60) Having regard to step c., “chromatographic separation” is to be understood as disclosed in step b., whereas the term “sample” means a real world sample, more specifically a real world lipid sample ranging, for example, from very simple vegetable oils to more complex fish oils, milk and milk-derived lipid samples such as, but not limited to, the following: vegetal oils, aquatic organisms (e.g. fish and algae) and derivatives thereof (e.g. fishmeal and fish oils), dairy products and biological samples as whole blood, plasma and cell membranes.

(61) As regards the determination of the retention index value of step d., it can be performed by means of Eq. 4 (reported here below), by using the odd-chain TAGs as reference homologous series according to the present invention (e.g., from C9C9C9 to C19C19C19).

(62) $\begin{array}{cc}\mathrm{LRI}=100\left[z+n\frac{t_{\mathrm{Ri}}-t_{\mathrm{Rz}}}{t_{R\left(z+n\right)}-t_{\mathrm{Rz}}}\right]& \mathrm{Eq}.4\end{array}$

(63) wherein,

(64) t.sub.Ri is the retention time of the unknown compound i, t.sub.Rz and t.sub.R(z+1) are the retention times of the TAGs which elute, respectively, immediately before and immediately after the unknown compound i, z is the PN associated with the compound eluted immediately before the unknown compound i, and n represents the difference in z units between the reference TAGs eluted immediately before and after the unknown compound i.

(65) As above anticipated, for the purpose of the present invention, z is equal to the PN of each reference TAG. Therefore, by way of example, z is 27 for C9C9C9, up to 57 for C19C19C19.

(66) Furthermore, since the difference in z units between two adjacent homologues is equal to 6, n is to be taken as 6 when using Eq. 4 for the calculation of LRIs.

(67) Step d. can further comprise the extrapolation of retention times by means of the calibration curve t.sub.R vs. PN established in step b (FIG. 2). In detail, given the linear relationship between t.sub.R and PN, it is possible to extrapolate the retention time of a reference compound not included in the employed homologue series (for example, PN minor than 27 or major than 57). More in particular, given the retention time of the unknown compound (step c.), it is possible to extrapolate the retention times and PN of both the reference TAG which elutes immediately before and after the unknown compound. Therefore, it will be possible to use such retention times and/or PN in equation 4 to calculate the LRI value of a compound eluted outside the chromatographic space covered by the homologous series.

(68) In another embodiment of the present invention, step d. can also be performed by means of the CromatoplusSpectra software, which automatically extrapolates retention indices for peaks eluted outside the chromatographic space covered by the homologous series. Such an extrapolation, again, is based on the direct proportionality/linearity between t.sub.R and PN, and can be also easily performed by each operator on an excel file.

(69) The identification step e. is performed by comparison of the calculated LRI value for the unknown compound with those previously obtained for listed compounds; if the calculated LRI of an unknown peak is the same of a reference, then a positive identification can be made.

(70) The term “listed compounds” comprises, but is not limited to, those established in step a. (i.e. LRI values of the reference TAGs). Said term in fact also comprises retention index values of all the compounds analysed by means of the present chromatographic method which have become part of the retention index library of the present invention (i.e. the LRI database comprising more than 209 TAGs and other simple and complex lipids such as, for example, free fatty acids, mono- and diacylglycerols (MAGs and DAGs, respectively), as well as sterols, carotenoids and waxes).

(71) In another embodiment of the present invention, the identification step e. can be performed by means of the CromatoplusSpectra software, which has been suitably developed to make automatic and fast the identification process only based on the match of the LRI value of an unknown compound with those of the LRI library previously established. The identification process by the new software consists of the following steps: peak integration in the chromatogram of the homologue series peaks to define their retention time; peak integration in the chromatogram of the real sample; loading in the LRI data processing window of the homologue series chromatogram previously integrated; loading of the LRI library (normally an excel file comprising just compound name and LRI value); LRI calculation for each unknown compound; match with the LRI library by one simple click; then a list of candidates is obtained depending on the LRI tolerance window set in the software.

(72) Regarding the chromatographic separation, in a preferred embodiment, the method comprises an UHPLC system equipped with two serially coupled narrow-bore columns (100×2.1 mm (L×ID)) wherein the stationary phase packing consists of sub-2 μm d.sub.p monodisperse particles and the mobile phase is acetonitrile(A)/2-propanol(B) under gradient conditions: 0-50% B in 105 min (held for 20 min). The flow rate is maintained at 400 μL/min with oven temperature of 35° C. As for the detection conditions, in a preferred embodiment, an ELSD detector is employed and the following ELSD parameters were applied: evaporative temperature 60° C., nebulizing gas (N.sub.2) pressure 270 kPa, detector gain 166<1 mV; sampling frequency: 10 Hz.

(73) The chromatogram of borage oil, menhaden oil and goat milk are reported in FIGS. 3A, 3B and 3C, respectively, in direct comparison with the reference homologue series chromatogram (FIG. 3D); an immediate identification (even by visual inspection) according to the partition number is shown, in order to highlight the inter sample repeatability in terms of retention time of each PN region, in spite of total different elution profiles. Moreover, the chromatograms of linseed oil, one of the most complex vegetable oil injected, and menhaden oil, the sample with the largest number of identified TAGs, are shown in FIG. 4, in order to point out the high chromatographic effectiveness of the present method.

(74) Table 1 below provides the list of abbreviations used to refer to the plurality of fatty acids in the analysed compounds.

(75) TABLE-US-00001 TABLE 1 Abbreviation Fatty Acid C7 heptanoic acid C9 nonanoic acid C11 undecanoic acid C13 tetradecanoic acid C15 pentadecanoic acid C17 heptadecanoic acid C19 nonadecanoic acid Bu butirric acid Cp caproic acid C caprylic acid Ca capric acid La lauric acid M myristic acid Pd pentadecanoic acid Ht hexadecatrienoic acid Po palmitoleic acid P palmitic acid St stearidonic acid Ln linolenic acid γLn gamma-linolenic acid L linoleic acid O oleic acid S stearic acid G gadoleic acid A arachidic acid Ep eicosapentaenoic acid Dh docosaesaenoico acid C22:1 docosenoic acid C24:1 tetracosenoic acid

(76) The identification based solely on the Retention Index System of the present invention is reliable in itself and leads to unequivocal identification (i.e. no retention indices superimposition occurs). However, sometimes, the obtained LRI value can positively match with more than one reference compound. This is the case of dilinoleyl-palmitoyl-glycerol (LLP), having a LRI of 4360±6, and dioleyl-linolenyl-glycerol (OOLn), having a LRI of 4358±2. Therefore, in order to avoid ambiguous identification, step e. can further comprise the inspection of the MS fragmentation pattern of each species, leading to the selection of the right candidate. FIG. 5 shows ESI-MS spectra of LLP (FIG. 5A) and OOLn (FIG. 5B) species, along with their fragment elucidation. In detail, each TAG shows a fragmentation profile characterized by the sodium adduct of the molecular species, diacylglycerols and monoacylglycerols fragments, leading to unequivocal identification. Going into the detail of this aspect of the invention, when an LC-MS system is employed, the same software is employed for the identification process, by adding one further step that is the interpretation of the MS spectrum that can be visualized by one simple click on the peak. In the near future, also a spectral library could be created and used as automatic identification tool in combination to the LRI database. In this case the software should be able to apply simultaneously both the identification filters to provide only the right candidate.

(77) Conversely, the method of the present invention can also be used to discriminate compounds having very similar or identical mass spectra. In summary, using retention indices in liquid chromatography coupled with MS, particularly EI-MS or API-MS, significantly increase the reliability of peak identifications by providing a second, independent experimental value. Therefore, it is possible to conclude that, according to the present invention, retention indices are not only essential parameters to perform automatic and fast searches in the LRI database, but they can also become complementary to MS allowing to discriminate between analytes with the same mass fragmentations, thus achieving a highly reliable identification.

(78) It is worth noting also that, to date, lipid profiling is commonly investigated by considering in a complementary way the GC profiles of fatty acid methyl esters (FAMEs) and LC profiles of intact lipids. In this way, the identification of intact lipids can be driven by the FA composition; in other words isobars (identical MS spectra) can be distinguished by considering the relative abundance of FAs, which are combined in the acylglycerols (MAGs, DAGs and TAGs) and phospholipids and present in sterol esters, carotenoid esters, waxes. Furthermore, for very simple samples, MS information could be not essential, being known FA composition and considering the typical elution profile of lipids under RP-LC. In fact, in RP-LC lipids are eluted according to the increasing PN, then a preliminary identification could be performed on the basis of the retention behavior. Nevertheless when considering complex samples, such as fish oils or milk and milk derived samples, many exceptions can be observed due to the unpredictable retention behaviour of highly polyunsaturated and short chain saturated fatty acids, which can be retained more weakly and elute in lower PN region of the chromatogram.

(79) To this regard, providing for a retention index value associated to each lipid, the method of the present invention represents a more reliable identification tool.

(80) As above anticipated, according to the present invention, there is further provided a library (i.e. database) of retention indices for listed compounds usable as effective identification tool in liquid chromatography.

(81) The retention index database was built by using the CromatoplusSpectra software (Chromaleont, Messina, Italy) which, since a linear relationship exists between retention index and chain length, is able to extrapolate LRI values for TAGs eluted earlier than C9C9C9 (down to triformin, C1C1C1) and subsequent to C19C19C19. Then, the analyses were processed by using the CromatoplusSpectra software, which allowed to directly match the LRI automatically calculated for each peak with those of the LRI database previously created. Since it returns a list of candidates as narrow as possible, it is possible to perform the search resting only upon the LRI database and/or without consider any spectral data. Moreover, the software was successfully tested on LC analysis carried out with different types of instrumentation (e.g., HPLC, UHPLC, nanoLC) coupled with different types of detectors (e.g., ELSD, API-MS, EI-MS etc.).

(82) As previously disclosed, the inventors were successful in finding an adequate reference mixture that covers all the analysis time and shows a regular elution profile under different experimental conditions.

(83) Therefore, due to both the regular retention behaviour of the TAGs homologous series and the robustness of the method of the present invention herein disclosed, it has been possible to build Table 2.

(84) Table 2 contains the list of all identified compounds, ordered according to their elution time, along with their PN and calculated LRI reported as confidence interval around the inter sample average. The absolute experimental maximum difference from the average (ΔLRI) is also reported and may be considered for the database search. The confidence interval was estimated at the 95% confidence level using a normal probability distribution, assuming that the uncertainty of the method is equal to the average of all the standard deviations obtained for each LRI value. Such assumption derives from the fact that the standard deviation distribution can be approximated to a Pearson chi-square (χ.sup.2) function. Therefore it is not related to the measured variable but only to its degrees of freedom, which correspond to the dimension or population of the same variable. In this specific situation the population (p) of the LRI value, indicated in bracket in Table 2 here below, corresponds to the number of samples (each in duplicate since two analyses were performed for each sample) comprising the specific compound.

(85) TABLE-US-00002 TABLE 2 List of identified compounds, along with their PN, LRI and ΔLRI value PN Compound Name LRI* ΔLRI 14 L (p = 10).sup.#  2451 ± 3.sup.a 6 16 O (p = 10).sup.#  2459 ± 3.sup.a 6 16 P (p = 10).sup.#  2519 ± 3.sup.a 8 18 S (p = 10).sup.#  2552 ± 3.sup.a 9 26 CaCC (p = 2)  2691 ± 6.sup.a 4 26 CaCaCp (p = 2)  2691 ± 6.sup.a 4 26 LaCCp (p = 2)  2691 ± 6.sup.a 4 28 LaCaCp (p = 2) 2800 ± 6 4 28 MCCp (p = 2) 2800 ± 6 4 28 OCBu (p = 6) 2816 ± 4 10 28 PCBu (p = 6) 2816 ± 4 10 28 LnO (p = 10).sup.$ 2953 ± 3 11 30 MCaCp (p = 10) 2969 ± 3 13 30 PCCp (p = 10) 3003 ± 3 13 30 OCaBu (p = 10) 3003 ± 3 13 30 PCaBu (p = 10) 3003 ± 3 13 30 MLaBu (p = 10) 3003 ± 3 13 30 OCpC (p = 2) 3014 ± 6 2 30 LO (p = 10).sup.$ 3123 ± 3 6 30 PL (p = 10).sup.$ 3153 ± 3 4 32 LaCaCa (p = 6) 3209 ± 4 5 32 MCaC (p = 4) 3223 ± 4 11 32 PLaBu (p = 8) 3241 ± 3 10 32 MLaCp (p = 4) 3241 ± 4 10 32 PCaCp (p = 10) 3243 ± 3 11 32 PCC (p = 2) 3249 ± 6 1 32 LLBu (p = 2) 3292 ± 6 0 32 PoPoBu (p = 2) 3292 ± 6 0 32 OO (p = 10).sup.$ 3336 ± 3 4 32 PO (p = 10).sup.$ 3362 ± 3 1 34 PCaC (p = 8) 3414 ± 3 6 34 MCaCa (p = 10) 3416 ± 3 10 34 OCaC (p = 2) 3426 ± 6 0 34 OLBu (p = 6) 3441 ± 4 4 34 LaLaCa (p = 2) 3443 ± 6 0 34 MLaC (p = 2) 3443 ± 6 0 34 OMBu (p = 10) 3454 ± 4 10 34 PMBu (p = 10) 3454 ± 4 10 30 EpEpEp (p = 2) 3463 ± 6 0 30 DhDhDh (p = 2) 3500 ± 6 0 34 SO (p = 6).sup.$ 3568 ± 3 7 36 OCaCa (p = 10) 3596 ± 4 8 36 OLaC (p = 6) 3597 ± 3 7 34 SP (p = 6).sup.$ 3599 ± 3 5 34 StStPo (p = 2) 3600 ± 6 1 34 EpPoHt (p = 2) 3600 ± 6 1 36 PPoCp (p = 8) 3604 ± 3 9 36 PCaCa (p = 10) 3604 ± 3 9 36 OLCp (p = 2) 3615 ± 6 1 36 OOBu (p = 8) 3620 ± 3 10 36 OMCp (p = 6) 3622 ± 4 8 34 EpStPo (p = 2) 3624 ± 6 1 36 MMC (p = 2) 3625 ± 6 1 36 MLaCa (p = 2) 3625 ± 6 1 36 LaLaLa (p = 2) 3625 ± 6 1 36 OPBu (p = 10) 3630 ± 3 10 36 PPBu (p = 10) 3630 ± 3 10 34 EpHtM (p = 2) 3649 ± 6 1 34 EpStM (p = 2) 3649 ± 6 1 36 PMCp (p = 2) 3655 ± 6 1 36 LnLnLn (p = 6) 3668 ± 4 5 34 EpEpPo (p = 2) 3669 ± 6 1 34 DhStPo (p = 2) 3669 ± 6 1 34 DhStM (p = 2) 3712 ± 6 1 34 EpEpM (p = 2) 3712 ± 6 1 36 γLnγLnγLn (p = 2) 3747 ± 6 1 34 DhEpPo (p = 2) 3755 ± 6 1 34 DhEpM (p = 2) 3755 ± 6 1 38 OOCp (p = 10) 3785 ± 3 8 38 OLaCa (p = 10) 3785 ± 3 8 38 OMC (p = 8) 3796 ± 3 8 36 EpLnPo (p = 2) 3799 ± 6 1 36 DhPHt (p = 2) 3807 ± 6 1 36 EpPHt (p = 2) 3807 ± 6 1 36 EpStP (p = 2) 3807 ± 6 1 38 PLaCa (p = 8) 3809 ± 3 8 38 MMCa (p = 6) 3810 ± 4 8 38 PoPoCa (p = 2) 3820 ± 6 1 38 PPoCo (p = 2) 3820 ± 6 1 38 PMC (p = 10) 3829 ± 3 7 38 LnLLn (p = 8) 3830 ± 3 3 36 EpEpO (p = 2) 3836 ± 6 1 38 OPCp (p = 10) 3842 ± 3 8 38 PPCp (p = 8) 3843 ± 4 7 36 EpEpP (p = 2) 3850 ± 6 1 36 DhEpO (p = 2) 3850 ± 6 1 36 DhEpP (p = 2) 3850 ± 6 1 38 γLnLγLn (p = 2) 3867 ± 6 1 38 PoHtM (p = 2) 3893 ± 6 1 38 EpPoPo (p = 2) 3893 ± 6 1 36 DhDhP (p = 2) 3893 ± 6 1 38 PoγLnγLn (p = 2) 3915 ± 6 3 38 DhPoPo (p = 2) 3937 ± 6 2 38 EpPoM (p = 2) 3937 ± 6 2 38 DhMM (p = 2) 3937 ± 6 2 40 OMCa (p = 8) 3987 ± 3 10 38 EpLnP (p = 2) 3987 ± 6 2 40 OOC (p = 8) 3989 ± 4 8 40 OPC (p = 10) 3991 ± 3 7 40 MMLa (p = 6) 3992 ± 4 7 40 OOCo (p = 6) 3992 ± 4 6 40 LLLn (p = 10) 3993 ± 3 6 40 SOCp (p = 6) 3993 ± 4 6 40 LLγLn (p = 2) 3999 ± 6 2 38 DhLnP (p = 2) 4004 ± 6 2 40 SMC (p = 10) 4009 ± 3 11 40 PPC (p = 10) 4009 ± 3 11 40 LnLnO (p = 6) 4011 ± 4 5 40 SSBu (p = 10) 4019 ± 4 6 40 LnPLn (p = 6) 4023 ± 4 5 38 DpDhP (p = 2) 4037 ± 6 2 38 DhEpS (p = 2) 4052 ± 6 2 40 γLnγLnO (p = 2) 4052 ± 6 2 40 γLnPγLn (p = 2) 4064 ± 6 2 40 LnMM (p = 2) 4094 ± 6 2 40 EpOPo (p = 2) 4094 ± 6 2 40 EpLP (p = 2) 4112 ± 6 2 40 DhOPo (p = 2) 4135 ± 6 2 40 DhOM (p = 2) 4148 ± 6 2 40 DhPPo (p = 2) 4148 ± 6 2 42 LLL (p = 24) 4160 ± 6 10 42 OOCa (p = 10) 4164 ± 2 8 42 OPCa (p = 8) 4164 ± 3 8 42 PoPoPo (p = 2) 4167 ± 4 2 42 PoPoM (p = 2) 4167 ± 6 2 42 PMLa (p = 2) 4175 ± 6 2 42 PPCa (p = 10) 4176 ± 3 9 42 SMCa (p = 8) 4178 ± 4 9 42 γLnLO (p = 2) 4181 ± 6 2 42 LnPPo (p = 2) 4186 ± 6 2 42 LPoPo 4186 ± 6 2 42 MMM (p = 2) 4190 ± 6 2 42 LnLO (p = 8) 4192 ± 3 11 42 SPC (p = 10) 4196 ± 3 10 42 SSCp (p = 10) 4196 ± 3 10 42 OOHt (p = 2) 4213 ± 4 3 42 SLnLn (p = 6) 4216 ± 4 4 42 LnLP (p = 6) 4217 ± 6 3 42 γLnLP (p = 2) 4221 ± 6 2 42 SγLnγLn (p = 2) 4221 ± 6 2 42 DhPP (p = 2) 4234 ± 6 3 42 DhOP (p = 2) 4246 ± 6 3 42 EpOO (p = 2) 4261 ± 6 3 44 OLPo (p = 2) 4286 ± 6 3 12 CE 18:3.sup.§ 4292 ± 6 1 12 CE 20:4.sup.§ 4292 ± 6 1 44 OPoM (p = 2) 4303 ± 6 3 44 PPoPo (p = 2) 4338 ± 6 3 44 LPM (p = 2) 4338 ± 6 3 44 GLγLn (p = 2) 4340 ± 6 3 44 LLO (p = 50) 4342 ± 1 10 44 OLM (p = 10) 4349 ± 3 5 44 LLP (p = 24) 4358 ± 2 7 44 OOLn (p = 14) 4360 ± 6 11 44 PPoM (p = 2) 4361 ± 2 3 44 OOγLn (p = 2) 4364 ± 6 3 44 OOLa (p = 10) 4364 ± 3 8 44 OPoPo (p = 8) 4367 ± 3 11 44 SLLn (p = 2) 4378 ± 6 3 44 EpSO (p = 2) 4380 ± 6 3 44 POLn (p = 4) 4383 ± 4 8 44 SPCa (p = 10) 4386 ± 3 9 44 PPLa (p = 10) 4386 ± 3 9 44 SMLa (p = 10) 4386 ± 3 9 44 PMM (p = 10) 4386 ± 3 9 44 SOCa (p = 8) 4387 ± 4 9 44 POγLn (p = 2) 4389 ± 6 3 44 ALnLn (p = 2) 4395 ± 6 3 44 PLnP (p = 2) 4395 ± 6 3 44 DhSO (p = 2) 4399 ± 6 3 44 EpSP (p = 2) 4416 ± 6 3 14 CE 18:2.sup.§ 4430 ± 6 1 44 PγLnP (p = 2) 4431 ± 6 3 44 DhSP (p = 2) 4433 ± 6 3 46 OOPo (p = 2) 4485 ± 6 3 46 C22:1LγLn (p = 2) 4501 ± 6 1 46 GLL (p = 2) 4502 ± 6 10 46 OOL (p = 42) 4516 ± 1 12 46 OOM (p = 10) 4526 ± 3 8 46 PLO (p = 46) 4539 ± 1 13 46 SLL (p = 8) 4548 ± 3 9 46 SOLa (p = 8) 4551 ± 3 15 46 PoPP (p = 2) 4552 ± 6 4 46 SOLn (p = 6) 4563 ± 4 7 46 PLP (p = 24) 4571 ± 2 14 46 DhSS (p = 2) 4577 ± 6 4 46 SMM (p = 10) 4580 ± 3 15 46 PPM (p = 10) 4580 ± 3 15 46 SLnP (p = 2) 4587 ± 6 4 46 SOγLn (p = 2) 4599 ± 6 4 46 SγLnP (p = 2) 4631 ± 6 4 47 OOPd (p = 2) 4690 ± 6 4 47 OPPd (p = 2) 4690 ± 6 4 48 C22:1LL (p = 2) 4690 ± 6 4 48 OLG (p = 2) 4703 ± 6 4 48 C24:1LγLn (p = 2) 4723 ± 6 9 48 OOO (p = 44) 4729 ± 1 14 16 CE 18:1.sup.§ 4734 ± 6 1 48 GLP (p = 2) 4740 ± 6 4 48 SLO (p = 16) 4746 ± 2 8 48 SLP (p = 2) 4750 ± 6 4 48 SOM (p = 10) 4754 ± 3 13 48 POO (p = 44) 4756 ± 2 13 48 POP (p = 26) 4776 ± 2 14 48 SMP (p = 10) 4780 ± 3 15 48 PPP (p = 10) 4784 ± 3 11 48 SPPo (p = 2) 4785 ± 6 4 50 C24:1LL (p = 2) 4881 ± 6 10 50 C22:1LO (p = 2) 4890 ± 6 4 50 GOO (p = 2) 4905 ± 6 7 50 GOP (p = 2) 4921 ± 6 8 50 SLS (p = 2) 4940 ± 6 4 50 SOO (p = 34) 4948 ± 2 14 50 SOP (p = 28) 4961 ± 3 15 50 SPP (p = 12) 4978 ± 3 8 52 C22:1γLnC22:1 (p = 2) 5069 ± 6 4 52 C24:1OL (p = 2) 5084 ± 6 12 52 C22:1OP (p = 2) 5084 ± 6 4 52 C22:1OO (p = 2) 5091 ± 6 4 52 SOS (p = 20) 5103 ± 2 15 52 SSP (p = 10) 5156 ± 3 15 54 C22:1OG (p = 2) 5238 ± 6 5 54 C22:1OS (p = 2) 5271 ± 6 5 54 SSS (p = 2) 5324 ± 6 5 *Intersample average; .sup.#Monoacylglycerol species; .sup.$diacylglycerol species; .sup.aextrapolated values; p: population; .sup.§CE = cholesterol esters for which PN coincides with the PN of the fatty acid bound to the cholesterol backbone.

(86) Table 2 also represents the retention indices library (i.e. database) of reference retention indices usable as effective identification tool in liquid chromatography. Said retention index library lists, but is not limited to, a total of 224 compounds identified according to their retention behaviour, obtained by means of the instant method based on the disclosed Retention Index System. In particular, the database can include all simple and complex lipids, such as, for example, free fatty acids, mono-, di- and triacylglycerols, as well as sterols, carotenoids and waxes.

(87) The length of the confidence interval was ≤12 units for all the compounds; as a consequence, a difference between subsequent TAGs (i.e. ΔLRI) higher than 12 LRI units avoids any peak mismatching. As above disclosed, in fact, the identification based only on the Retention Index System of the present invention is reliable in itself and can lead to unequivocal identification (i.e. no retention indices superimposition occurs).

(88) Within this context the chromatographic resolution acquires special importance; it is strongly related to the so-called separation number (SN), which express the number of peaks that can be resolved between two consecutive TAGs of the homologue series (z and z+6, respectively):

(89) $\begin{array}{cc}SN=\frac{t_{R\left(z+6\right)}-t_{Rz}}{w_{h\left(z+6\right)}+w_{hz}}-1& \mathrm{Eq}.5\end{array}$

(90) wherein,

(91) t.sub.R(z+6) and t.sub.Rz are the retention times of two consecutive TAGs of the homologue series and w.sub.h(z+6) and w.sub.hz are the corresponding peak width at half height.

(92) By applying Eq. 5 to the region z=39 and z=45 (arbitrarily chosen, since a great number of TAGs are contained in this region), SN is equal to 25 that means that 25 peaks can be satisfactorily resolved between C13C13C13 and C15C15C15 standard compounds. As a consequence the minimum retention index difference (i.e. δ) between two adjacent peaks must be:
δ=600/(SN+1)

(93) equal in the present situation to 23, which largely overcomes the obtained ΔLRI above described. In order to achieve accurate library search results, the experimental LRI interval length (i.e. ΔLRI*2) must be also minor than δ value (ΔLRI*2<δ). For instance, within the 3900-4500 LRI region, for which δ is equal to 23, ΔLRI should not exceed a value of 11.5 (δ/2, hereinafter also referred as maximum tolerable variability) that means that a larger variation will cause an overlapping between adjacent peaks. Therefore, if two adjacent peaks have a LRI difference equal to or higher than 23 and an index variability lower than ±11.5, no mismatching or misidentification occurs. Accordingly, SN and δ calculations were carried out for all subsequent reference compound peaks, finding that, for each region, δ was equal or even higher than the experimental LRI interval length (δ≥ΔLRI*2). In particular, a total SN of 122 and an average δ of 30 units were obtained, meaning that ΔLRI values minor than 15 normally ensure a reliable identification (δ/2≥ΔLRI), thus demonstrating the high accuracy of the method of the present invention. For evidence, Table 3 is provided here below.

(94) TABLE-US-00003 TABLE 3 SN between adjacent Compound LRI ± ΔLRI compounds C7C7C7 2100 CaCC  2691 ± 4* 12 CaCaCp  2691 ± 4* C9C9C9 2700 OCBu  2816 ± 10 21 PCBu  2816 ± 10 LaCaCa 3209 ± 5 MCaC  3223 ± 11 PlaBu  3241 ± 10 MlaCp  3241 ± 10 PcaCp  3243 ± 11 C11C11C11 3300 PcaC 3414 ± 6 29 McaCa  3416 ± 10 OLBu 3441 ± 4 EpEpEp 3463 ± 1 DhDhDh 3500 ± 1 OlaC 3597 ± 7 StStPo 3600 ± 1 EpPoHt 3600 ± 1 PpoCp 3604 ± 9 OOBu  3620 ± 10 OMCp 3622 ± 8 EpStPo 3624 ± 1 EpHtM 3649 ± 1 EpStM 3649 ± 1 LnLnLn 3668 ± 5 EpEpPo 3669 ± 1 DhStPo 3669 ± 1 DhStM 3712 ± 1 EpEpM 3712 ± 1 DhEpPo 3755 ± 1 DhEpM 3755 ± 1 OMC 3796 ± 8 EpLnPo 3799 ± 1 DhPHt 3807 ± 1 EpPHt 3807 ± 1 EpStP 3807 ± 1 PlaCa 3809 ± 8 MMCa 3810 ± 8 LnLLn 3830 ± 3 EpEpO 3836 ± 1 EpEpP 3850 ± 1 DhEpP 3850 ± 1 EpPoPo 3893 ± 1 DhDhP 3893 ± 1 C13C13C13 3900 EpPoM 3937 ± 2 25 DhMM 3937 ± 2 DhPoPo 3937 ± 2 EpLnP 3987 ± 2 OOC 3989 ± 8 LLLn 3993 ± 6 SOCp 3993 ± 6 DhLnP 4004 ± 2 LnLnO 4011 ± 4 LnPLn 4023 ± 5 DpDhP 4037 ± 2 DhEpS 4052 ± 2 LnMM 4094 ± 2 EpOPo 4094 ± 2 EpLP 4112 ± 2 DhOPo 4135 ± 2 DhOM 4148 ± 2 DhPPo 4148 ± 2 LLL  4160 ± 10 OOCa 4164 ± 8 OPCa 4164 ± 8 PoPoM 4167 ± 2 PPCa 4176 ± 9 SMCa 4178 ± 9 LnPPo 4186 ± 2 LpoPo 4186 ± 2 LnLO  4192 ± 11 OOHt 4213 ± 3 LnLP 4216 ± 6 SLnLn 4216 ± 3 DhPP 4234 ± 3 DhOP 4246 ± 3 EpOO 4261 ± 3 OLPo 4286 ± 3 OpoM 4303 ± 3 PpoPo 4338 ± 3 LPM 4338 ± 3 LLO  4342 ± 10 OLM 4349 ± 5 LLP 4358 ± 7 OOLn  4360 ± 11 PpoM 4361 ± 3 OOLa 4364 ± 8 OpoPo  4367 ± 11 EpSO 4380 ± 3 POLn 4383 ± 8 PLnP 4395 ± 3 DhSO 4399 ± 3 EpSP 4416 ± 3 DhSP 4433 ± 3 OOPo 4485 ± 3 C15C15C15 4500 OOL  4516 ± 12 19 OOM 4526 ± 8 PLO  4539 ± 13 SLL 4548 ± 9 PoPP 4552 ± 4 SOLn 4563 ± 7 PLP  4571 ± 14 DhSS 4577 ± 4 OOPd 4690 ± 4 OPPd 4690 ± 4 OOO  4729 ± 14 SLO 4746 ± 8 SOM  4754 ± 13 POO  4756 ± 13 POP  4776 ± 14 PPP  4784 ± 11 SPPo 4785 ± 4 SOO  4948 ± 14 SOP  4961 ± 15 SPP 4978 ± 8 C17C17C17 5100 SOS  5103 ± 15 16 SSP  5156 ± 15 TOTALE 122 *Extrapolated values

(95) It shows most common identified compounds along with their LRI value, expressed as mean value±experimental variability, and SN value for each pair of adjacent homologues, from which the maximum tolerable variability can be deduced. The experimental variability is lower than the maximum tolerable variability for all the regions of the chromatogram, demonstrating the effectiveness of the instant method. The homologous series chromatogram is shown in FIG. 6 together with both SN and ΔLRI values for each pair of adjacent homologues.

(96) A similar consideration can arise taking into account the average peak width (w.sub.b) of the analytes. In fact, being w.sub.b=0.8 min (48 s) and 1 LRI=25 min/600=0.042 min (2.52 s), where 25 minutes is the difference time between C13C13C13 and C15C15C15 retention times, a LRI variability (ΔLRI) of ±15 units corresponding to 0.63 min (37.8 s) results minor than w.sub.b (i.e. ΔLRI<w.sub.b). In other words, the retention indices obtained have determination repeatability minor than the base peak width of the unknown compounds.

(97) Such a result again proves that each LRI corresponds only to the compound which occupies a specific position in the chromatogram in terms of retention time, pinpointing the reliability of a potential identification based only on the retention index system of the present invention.

(98) The novel Retention Index System of the present invention and the LRI database based thereon has succeed in the identification of a multitude of different real world lipid samples such as, but not limited to, the following: vegetal oils, aquatic organisms (e.g. fish and algae) and derivatives thereof (e.g. fishmeal and fish oils), dairy products and biological samples as whole blood, plasma and cell membranes.

DESCRIPTION OF PREFERRED EMBODIMENTS

(99) The embodiments herein described have illustrative purpose only and are not meant to limit the scope of the claimed invention. Moreover, modifications and alterations of the embodiments below, obvious to a man skilled in the art, are intended to be encompassed by the attached claims.

Samples Preparation

(100) For the purpose of the present invention, any lipid sample is a suitable sample. In particular, vegetable oils, fish oils, milk fats, milk-derived fats, derivatives and combinations thereof are all suitable samples. In some embodiments, vegetable oils (e.g., peanut, corn, soybean, rapeseed, grapeseed, almond, hazelnut, sunflower, linseed, chia, sacha inchi, borage, castor, olive), fish oils (e.g., menhaden, cod liver, sea bream, tuna fish, fish integrator, shrimps) and milk and milk derived fats (e.g., goat, cow, whole biological milk, buffalo mozzarella, butter) were selected. Lipid samples preparation was performed according to any of the suitable methods known in the art for the extraction of the lipid phase (e.g., lipid extraction using organic solvents or by the Folch and Bligh&Dyer procedures, for milk and fish samples respectively). Usually, the injection is performed at a concentration of 1-10 mg/ml.

LRI Calculation and Data Analysis

(101) LRI were calculated according to Eq. 4, by using the odd-chain TAGs as reference homologous series according to the present invention (e.g., from C9C9C9 to C19C19C19) and by assigning to z a value equal to the partition number (PN) of each reference TAG that is related to the carbon chain length (CN) and double bond number (DB) accordingly to the following equation:
PN=CN−2DB

(102) Therefore z is 27 for C9C9C9, up to 57 for C19C19C19, and n is equal to 6.

(103) The LRI database was built by using the CromatoplusSpectra software (Chromaleont, Messina, Italy), that was also able to extrapolate LRI values for compounds eluted earlier than C9C9C9 and subsequent to C19C19C19. Then analyses were processed by using the CromatoplusSpectra software, which allowed to directly match the LRI automatically calculated for each peak with those of the LRI database previously created.

Evidence of the Method Independency: Repeatability Studies

(104) The inventors of the present invention have successfully established a novel Retention Index System such that the chromatographic method based thereon is able not only to provide maximum resolution for a wide range of lipid samples but it is also independent from the selected chromatographic conditions and, as a consequence, repeatable both at intra- and inter-laboratory level.

(105) As regards the method repeatability, the inventors of the present invention have also performed a systematic study in order to assess the LRI stability under a wide range of experimental conditions. In particular, all the analytical parameters playing a role in the chromatographic separation such as, for example, column dimension, stationary phase packing, flow rate, oven temperature, gradient and mobile phase composition were accurately evaluated by injecting three vegetable oils (borage, linseed and olive oils) comprising a total of 54 TAGs. This repeatability study is not limited to the selected oils, but rather is representative for all the lipid samples for which the present chromatographic method is suitable.

(106) Regarding the columns, two different stationary phase packing were compared: monodisperse sub-2 μm (particle size 1.9 μm) and partially porous (particle size 2.7 and 2.0 μm) particles. In addition, different column dimensions were used: 100×2.1 mm, L×ID, operated at 400 μL/min and 100×4.6 mm, L×ID, operated at 1.8 mL/min in order to evaluate the internal diameter effect. The stationary phase bed length was also changed by considering two serially coupled columns. The results are reported in Table 4 below, which contains the list of TAGs along with their PN and LRI value on each column. The inter-sample average and ΔLRI are also provided. The experiment shows that the method of the present invention is independent from both column dimensions and stationary phase packing since a variability (ΔLRI)≤15 was obtained.

(107) TABLE-US-00004 TABLE 4 LRI 2 Titan 1 Titan 100 mm × 100 mm × 2.1 mm ID, 2.1 mm ID, Compound 1.9 μm dp 1.9 μm dp PN Name (monodisperse).sup.a (monodisperse).sup.b 36 LnLnLn 3668 3673 36 γLnγLnγLn 3747 3738 38 LnLLn 3830 3840 38 γLnLγLn 3867 3886 38 PoγLnγLn 3915 3914 40 LLLn 3993 4002 40 LLγLn 3999 4022 40 LnLnO 4011 4012 40 LnPLn 4023 — 40 γLnγLnO 4052 4071 40 γLnPγLn 4086 4084 42 LLL 4160 4165 42 γLnLO 4181 4204 42 LnLO 4192 4188 42 SLnLn 4216 4206 42 LnLP 4217 4227 42 γLnLP 4221 4223 42 SγLnγLn 4221 4223 44 LLO 4342 4348 44 LLP 4358 4364 44 OOLn 4360 4367 44 OOγLn 4364 4382 44 POLn 4383 4387 44 POγLn 4389 4408 44 PγLnP 4431 4428 46 GLL 4502 4511 46 OOL 4516 4525 46 PLO 4539 4551 46 SLL 4548 — 46 SOLn 4563 4574 46 PLP 4571 4574 46 SOγLn 4599 4588 46 SγLnP 4631 4625 48 C22:1LL 4690 4701 48 OLG 4703 4711 48 C24:1LγLn 4723 4732 48 OOO 4729 4734 48 GLP 4740 — 48 POO 4756 4762 48 POP 4776 4781 50 C24:1LL 4881 4880 50 C22:1LO 4890 4893 50 GOO 4905 4908 50 GOP 4921 4923 50 SLS 4940 — 50 SOO 4948 4949 50 SOP 4961 — 50 SPP 4978 — 52 C22:1γLnC22:1 5069 5065 52 C24:1OL 5084 5081 52 C22:1OO 5091 5095 52 SOS 5103 5103 54 C22:1OG 5238 5251 54 C22:1OS 5271 5282 LRI 1 Ascentis 1 Ascentis Express Express 100 mm × 100 mm × 2.1 mm ID, 4.6 mm ID, 2.7 μm dp 2.7 μm dp Compound (partially (partially PN Name porous).sup.b porous).sup.c 36 LnLnLn 3667 3669 36 γLnγLnγLn 3740 3732 38 LnLLn 3842 3842 38 γLnLγLn 3893 3889 38 PoγLnγLn 3917 — 40 LLLn 4006 4005 40 LLγLn 4022 4022 40 LnLnO 4019 4015 40 LnPLn 4042 4040 40 γLnγLnO 4078 4072 40 γLnPγLn 4094 4088 42 LLL 4170 4160 42 γLnLO 4205 4201 42 LnLO 4193 4186 42 SLnLn 4212 — 42 LnLP 4236 4229 42 γLnLP 4226 4221 42 SγLnγLn 4226 4221 44 LLO 4348 4339 44 LLP 4365 4360 44 OOLn 4370 4365 44 OOγLn 4386 4379 44 POLn 4393 4389 44 POγLn 4413 4407 44 PγLnP 4435 4431 46 GLL 4517 4511 46 OOL 4531 4522 46 PLO 4557 4548 46 SLL 4575 4575 46 SOLn 4588 — 46 PLP 4594 4584 46 SOγLn 4610 4602 46 SγLnP 4638 4629 48 C22:1LL 4706 4699 48 OLG 4711 4710 48 C24:1LγLn 4717 4723 48 OOO 4740 4732 48 GLP 4732 4740 48 POO 4769 4763 48 POP 4787 4779 50 C24:1LL 4891 4882 50 C22:1LO 4896 4891 50 GOO 4910 4904 50 GOP 4930 4919 50 SLS — — 50 SOO 4947 4938 50 SOP 4945 4943 50 SPP 4978 4972 52 C22:1γLnC22:1 5071 — 52 C24:1OL 5080 5078 52 C22:1OO 5097 5088 52 SOS 5106 5106 54 C22:1OG 5243 5228 54 C22:1OS 5271 5267 LRI 1 Ascentis Express 100 mm × 2.1 mm ID, 2 μm dp Compound (partially TOTAL PN Name porous).sup.b AVERAGE 36 LnLnLn 3654 3665 36 γLnγLnγLn 3733 3738 38 LnLLn 3823 3835 38 γLnLγLn 3874 3882 38 PoγLnγLn 3913 3915 40 LLLn 3989 3999 40 LLγLn 4009 4015 40 LnLnO 4002 4012 40 LnPLn 4031 4034 40 γLnγLnO 4063 4067 40 γLnPγLn 4081 4086 42 LLL 4151 4161 42 γLnLO 4189 4196 42 LnLO 4175 4187 42 SLnLn 4198 4208 42 LnLP 4223 4226 42 γLnLP 4215 4221 42 SγLnγLn 4215 4221 44 LLO 4331 4341 44 LLP 4352 4360 44 OOLn 4356 4364 44 OOγLn 4367 4375 44 POLn 4382 4387 44 POγLn 4402 4403 44 PγLnP 4430 4431 46 GLL 4498 4508 46 OOL 4515 4522 46 PLO 4546 4548 46 SLL 4550 4562 46 SOLn 4577 4575 46 PLP 4579 4580 46 SOγLn 4598 4599 46 SγLnP 4634 4631 48 C22:1LL 4692 4698 48 OLG 4703 4708 48 C24:1LγLn 4719 4723 48 OOO 4727 4732 48 GLP 4735 4737 48 POO 4761 4762 48 POP 4783 4781 50 C24:1LL 4873 4881 50 C22:1LO 4883 4890 50 GOO 4898 4905 50 GOP 4914 4921 50 SLS 4940 4940 50 SOO 4938 4944 50 SOP 4951 4950 50 SPP 4975 4976 52 C22:1γLnC22:1 5052 5064 52 C24:1OL 5065 5077 52 C22:1OO 5085 5091 52 SOS 5099 5103 54 C22:1OG 5230 5238 54 C22:1OS 5265 5271 Compound PN Name ΔLRI 36 LnLnLn 11 36 γLnγLnγLn 9 38 LnLLn 13 38 γLnLγLn 14 38 PoγLnγLn 3 40 LLLn 10 40 LLγLn 15 40 LnLnO 10 40 LnPLn 11 40 γLnγLnO 15 40 γLnPγLn 7 42 LLL 10 42 γLnLO 14 42 LnLO 11 42 SLnLn 10 42 LnLP 10 42 γLnLP 6 42 SγLnγLn 6 44 LLO 10 44 LLP 8 44 OOLn 8 44 OOγLn 12 44 POLn 6 44 POγLn 15 44 PγLnP 4 46 GLL 10 46 OOL 9 46 PLO 9 46 SLL 14 46 SOLn 13 46 PLP 14 46 SOγLn 11 46 SγLnP 6 48 C22:1LL 8 48 OLG 5 48 C24:1LγLn 9 48 OOO 7 48 GLP 5 48 POO 7 48 POP 6 50 C24:1LL 10 50 C22:1LO 8 50 GOO 7 50 GOP 8 50 SLS 0 50 SOO 6 50 SOP 11 50 SPP 4 52 C22:1γLnC22:1 12 52 C24:1OL 12 52 C22:1OO 7 52 SOS 4 54 C22:1OG 13 54 C22:1OS 11 .sup.aflow rate 400 μL/min, gradient: 0-105 min, 0-50% B (hold 20 min); .sup.bflow rate 400 μL/min, gradient: 0-52.5 min, 0-50% B (hold 10 min); .sup.cflow rate 1.8 mL/min, gradient: 0-52.5 min, 0-50% B (hold 10 min). Oven temperature was 35° C. in all the analyses.

(108) It is worth noting also that, despite previous works taught the use of wider-bore columns (namely 4.6 mm I.D or greater), the method of the present invention using narrow-bore columns (2.1 mm I.D) and operating at a smaller flow rate allows a more straightforward coupling with MS detectors, if any, in addition to a significant solvent saving.

(109) The partially porous 100×2.1 mm, L×ID, 2.0 μm d.sub.p, column was then selected for the evaluation of the effect of the other parameters.

(110) Three different flow rates were then compared: 300 μL/min, 400 μL/min, and 500 μL/min in gradient mode (i.e. 0-52.5 min, 0-50% B (hold 10 min)) at 35° C. The below Table 5 shows the LRI values calculated at different flow rates along with their average ΔRLI, highlighting a perfect agreement between all the conditions (ΔLRI≤10) and leading to conclude that the method of the present invention is independent even from the flow rate.

(111) TABLE-US-00005 TABLE 5 LRI Compound 300 400 500 TOTAL PN Name μL/min μL/min μL/min AVERAGE ΔLRI 36 LnLnLn 3651 3647 3643 3647 4 36 γLnγLnγLn 3733 3733 3732 3732 1 38 LnLLn 3820 3823 3826 3823 3 38 γLnLγLn 3869 3874 3880 3874 6 38 PoγLnγLn 3901 3913 3918 3910 10 40 LLLn 3982 3989 3995 3989 7 40 LnLnO 3994 4002 4007 4001 7 40 LLγLn 4003 4009 4013 4008 5 40 LnPLn 4025 4031 4036 4030 6 40 γLnγLnO 4054 4063 4067 4061 7 40 γLnPγLn 4075 4081 4084 4080 5 42 LLL 4144 4151 4157 4150 7 42 LnLO 4167 4175 4181 4175 8 42 γLnLO 4182 4189 4193 4188 6 42 SLnLn 4190 4198 4203 4197 7 42 γLnLP 4208 4215 4217 4213 5 42 SγLnγLn 4208 4215 4217 4213 5 42 LnLP 4215 4223 4229 4222 7 44 LLO 4323 4331 4337 4330 7 44 LLP 4345 4352 4357 4351 7 44 OOLn 4348 4356 4362 4355 8 44 OOγLn 4357 4367 4373 4366 9 44 POLn 4375 4382 4388 4382 7 44 POγLn 4393 4402 4406 4400 7 44 PγLnP 4424 4430 4434 4429 5 46 GLL 4492 4498 4505 4498 6 46 OOL 4510 4515 4521 4515 6 46 PLO 4541 4546 4549 4545 5 46 SLL 4546 4550 — 4548 2 46 SOLn 4571 4577 4580 4576 5 46 PLP 4577 4579 4585 4580 5 46 SOγLn 4591 4598 4603 4597 6 46 SγLnP 4625 4634 4634 4631 6 48 C22:1LL 4682 4692 4695 4690 8 48 OLG 4695 4703 4706 4701 7 48 C24:1LγLn 4713 4719 4721 4718 5 48 OOO 4720 4727 4729 4726 6 48 GLP 4725 4735 4738 4733 8 48 POO 4755 4761 4762 4759 4 48 POP 4778 4783 4784 4781 4 50 C24:1LL 4864 4873 4877 4871 7 50 C22:1LO 4874 4883 4885 4880 7 50 GOO 4890 4898 4899 4895 6 50 GOP 4909 4914 4917 4913 5 50 SOO 4933 4938 4938 4936 4 50 SLS 4936 4940 4941 4939 3 50 SOP 4944 4951 4954 4949 6 50 SPP 4968 4975 4973 4972 4 52 C22:1γLnC22:1 5044 5052 5053 5050 6 52 C24:1OL 5054 5065 5068 5062 8 52 C22:1OO 5079 5085 5085 5083 4 52 SOS 5099 5099 5097 5098 1 54 C22:1OG 5222 5230 5229 5227 5 54 C22:1OS 5257 5265 5264 5262 5

(112) Regarding the gradient of the elution phase, three different gradient steepness (% B/min) were compared, namely, 0.95% B/min, 1.4% B/min, and 0.70% B/min. Table 6 below shows the LRI values calculated at different gradient steepness along with their average ΔRLI demonstrating again that the method is not significantly affected by this change (ΔLRI≤15).

(113) TABLE-US-00006 TABLE 6 LRI Compound 0.95% 1.4% 0.70% TOTAL PN Name B/min B/min B/min AVERAGE ΔLRI 36 LnLnLn 3647 3647 3653 3649 4 36 γLnγLnγLn 3733 3730 3737 3733 4 38 LnLLn 3823 3815 3836 3825 12 38 γLnLγLn 3874 3865 3891 3877 15 38 PoγLnγLn 3913 3899 — 3906 7 40 LLLn 3989 3978 4002 3989 12 40 LnLnO 4002 3991 4012 4002 11 40 LLγLn 4009 3998 4020 4009 11 40 LnPLn 4031 4021 4040 4030 10 40 γLnγLnO 4063 4051 4073 4062 11 40 γLnPγLn 4081 4071 4091 4081 10 42 LLL 4151 4139 4159 4150 11 42 LnLO 4175 4163 4185 4174 12 42 γLnLO 4189 4177 4198 4188 11 42 SLnLn 4198 4186 4207 4197 11 42 γLnLP 4215 4204 4223 4214 10 42 SγLnγLn 4215 4204 4223 4214 10 42 LnLP 4223 4210 4234 4222 12 44 LLO 4331 4317 4342 4330 13 44 OOLn 4356 4343 4367 4355 13 44 LLP 4352 4340 4362 4351 12 44 OOγLn 4367 4353 4379 4366 13 44 POLn 4382 4369 4394 4381 13 44 POγLn 4402 4389 4411 4400 12 44 PγLnP 4430 4422 4435 4429 7 46 GLL 4498 4486 4512 4499 13 46 OOL 4515 4504 4527 4515 12 46 PLO 4546 4535 4554 4545 10 46 SLL 4550 4569 4579 4566 16 46 SOLn 4577 4564 4583 4575 11 46 PLP 4579 4576 4598 4584 14 46 SOγLn 4598 4618 4604 4607 11 46 SγLnP 4634 — 4636 4635 1 48 C22:1LL 4692 4687 4700 4693 7 48 OLG 4703 4711 4710 4708 5 48 C24:1LγLn 4719 4720 4726 4722 4 48 OOO 4727 4727 4733 4729 4 48 GLP 4735 4735 4744 4738 6 48 POO 4761 4779 4766 4769 11 48 POP 4783 4776 — 4779 4 50 C24:1LL 4873 4866 — 4869 4 50 C22:1LO 4883 4885 — 4884 1 50 GOO 4898 4903 — 4900 3 50 GOP 4914 4916 — 4915 1 50 SOO 4938 4932 4947 4939 8 50 SLS 4940 4944 — 4942 2 50 SOP 4951 4966 — 4959 8 50 SPP 4975 — — 4975 0 52 C22:1γLnC22:1 5052 5054 — 5053 1 52 C24:1OL 5065 5074 — 5069 4 52 C22:1OO 5085 5087 — 5086 1 52 SOS 5099 — — 5099 0 54 C22:1OG 5230 — — 5230 0 54 C22:1OS 5265 — — 5265 0

(114) Regarding the temperature, three different oven temperature were compared in the range 30-40° C., namely, 30° C., 35° C. and 40° C., leading to the conclusion that a variation of 10° C. can affect the LRI value, while 5° C. difference could not have a strong influence on the chromatographic method. Table 7 below shows both the average between 30 and 35° C. and between 35 and 40° C., along with their ΔLRI values, showing once again that the method of the present invention is reliable and repeatable even in case of a differential temperature (i.e. up to 5° C.) with a maximum ΔLRI equal to 13.

(115) TABLE-US-00007 TABLE 7 Compound LRI PN Name 30° C. 35° C. 40° C. 36 LnLnLn 3651 3647 3651 36 γLnγLnγLn 3755 3733 3724 38 LnLLn 3819 3823 3835 38 γLnLγLn 3866 3874 3891 38 PoγLnγLn 3897 3913 3927 40 LLLn 3982 3989 4001 40 LnLnO 3993 4002 4013 40 LLγLn 3998 4009 4020 40 LnPLn 4028 4031 4035 40 γLnγLnO 4053 4063 4073 40 γLnPγLn 4075 4081 4088 42 LLL 4140 4151 4162 42 LnLO 4165 4175 4186 42 γLnLO 4178 4189 4198 42 SLnLn 4192 4198 4205 42 γLnLP 4206 4215 4220 42 SγLnγLn 4206 4215 4220 42 LnLP 4216 4223 4231 44 LLO 4320 4331 4342 44 OOLn 4349 4356 4366 44 LLP 4346 4352 4360 44 OOγLn 4365 4367 4382 44 POLn 4378 4382 4391 44 POγLn 4395 4402 4410 44 PγLnP 4429 4430 4432 46 GLL 4485 4498 4516 46 OOL 4502 4515 4528 46 PLO 4535 4546 4554 46 SLL 4550 4550 4575 46 SOLn 4563 4577 4584 46 PLP 4573 4579 4588 46 SOγLn 4586 4598 4607 46 SγLnP 4624 4634 4633 48 C22:1LL 4670 4692 4700 48 OLG 4684 4703 4707 48 C24:1LγLn 4704 4719 4722 48 OOO 4712 4727 4730 48 GLP 4717 4735 4739 48 POO 4749 4761 4760 48 POP 4774 4783 4779 50 C24:1LL 4849 4873 4889 50 C22:1LO 4860 4883 4893 50 GOO 4878 4898 4907 50 GOP 4897 4914 4923 50 SLS 4915 4940 4945 50 SOO 4913 4938 4944 50 SOP 4927 4951 4972 50 SPP 4968 4975 4973 52 C22:1γLnC22:1 5034 5052 5067 52 C24:1OL 5049 5065 5077 52 C22:1OO 5069 5085 5095 52 SOS 5086 5099 5103 54 C22:1OG 5212 5230 5239 54 C22:1OS 5248 5265 5269 AVERAGE ΔLRI Compound (35° C./ (35° C./ PN Name ΔLRI 40° C.) 40° C.) 36 LnLnLn 11 3649 2 36 γLnγLnγLn 27 3728 5 38 LnLLn 29 3829 6 38 γLnLγLn 31 3882 9 38 PoγLnγLn 27 3920 7 40 LLLn 23 3995 6 40 LnLnO 22 4007 5 40 LLγLn 24 4014 6 40 LnPLn 4 4033 2 40 γLnγLnO 22 4068 5 40 γLnPγLn 14 4084 3 42 LLL 21 4156 5 42 LnLO 22 4180 5 42 γLnLO 22 4193 5 42 SLnLn 11 4201 4 42 γLnLP 11 4217 2 42 SγLnγLn 11 4217 2 42 LnLP 17 4227 4 44 LLO 26 4337 5 44 OOLn 22 4361 5 44 LLP 18 4356 4 44 OOγLn 29 4374 7 44 POLn 16 4386 4 44 POγLn 21 4406 4 44 PγLnP 2 4431 1 46 GLL 16 4507 9 46 OOL 34 4521 7 46 PLO 25 4550 4 46 SLL 13 4562 13 46 SOLn 28 4580 3 46 PLP 12 4583 4 46 SOγLn 11 4602 4 46 SγLnP 17 4633 1 48 C22:1LL 40 4696 4 48 OLG 34 4705 2 48 C24:1LγLn 11 4720 1 48 OOO 25 4728 1 48 GLP 26 4737 2 48 POO 15 4760 0 48 POP 6 4781 2 50 C24:1LL 39 4881 8 50 C22:1LO 278 4888 5 50 GOO 16 4902 5 50 GOP 28 4919 5 50 SLS 18 4942 3 50 SOO 27 4941 3 50 SOP 32 4962 11 50 SPP 4 4974 1 52 C22:1γLnC22:1 40 5060 8 52 C24:1OL 42 5071 6 52 C22:1OO 31 5090 5 52 SOS 28 5101 2 54 C22:1OG 32 5234 5 54 C22:1OS 16 5267 2 AVERAGE ΔLRI Compound (30° C./ (30° C./ PN Name 35° C.) 35° C.) 36 LnLnLn 3649 2 36 γLnγLnγLn 3744 0 38 LnLLn 3821 2 38 γLnLγLn 3870 4 38 PoγLnγLn 3905 8 40 LLLn 3986 4 40 LnLnO 3998 5 40 LLγLn 4003 5 40 LnPLn 4030 2 40 γLnγLnO 4058 5 40 γLnPγLn 4078 3 42 LLL 4146 5 42 LnLO 4170 5 42 γLnLO 4183 5 42 SLnLn 4195 3 42 γLnLP 4211 5 42 SγLnγLn 4211 5 42 LnLP 4219 4 44 LLO 4325 6 44 OOLn 4352 4 44 LLP 4349 3 44 OOγLn 4366 1 44 POLn 4380 2 44 POγLn 4398 3 44 PγLnP 4429 0 46 GLL 4492 7 46 OOL 4508 7 46 PLO 4541 6 46 SLL 4550 0 46 SOLn 4570 7 46 PLP 4576 3 46 SOγLn 4592 6 46 SγLnP 4629 5 48 C22:1LL 4681 11 48 OLG 4693 10 48 C24:1LγLn 4711 8 48 OOO 4720 8 48 GLP 4726 9 48 POO 4755 6 48 POP 4778 4 50 C24:1LL 4861 0 50 C22:1LO 4871 11 50 GOO 4888 10 50 GOP 4906 9 50 SLS 4927 13 50 SOO 4926 6 50 SOP 4939 0 50 SPP 4971 4 52 C22:1γLnC22:1 5043 9 52 C24:1OL 5057 8 52 C22:1OO 5077 8 52 SOS 5092 7 54 C22:1OG 5221 9 54 C22:1OS 5256 8

(116) Therefore it was successfully demonstrated that the method of the present invention is almost independent from all the variables involved in the chromatographic apparatus.

(117) Retention indices repeatability studies were carried out in order to assess and demonstrate also the robustness of the novel identification method and to make the LRI database usable both at intra- and inter-laboratory levels.

Reliability Studies by MS Analyses

(118) Among the 224 compounds listed in Table 2, 54 TAGs representative of vegetable oils were positively confirmed by HPLC-ESI-MS analyses, while UHPLC-ESI-MS analyses will be necessary to confirm identification for fish and milk samples. The obtained results demonstrates the great reliability of the method of the present invention.

(119) Nevertheless, if completely co-eluted, different compounds might have the same LRI. However, it is not probable that they present also the same mass spectrum. To this purpose, HPLC-ESI-MS analyses of borage, linseed and olive oils were carried out, providing the confirmation of the preliminary identification by UHPLC-ELSD method for 54 TAGs common in vegetable oils. Also in this case, a comparison between the LRI values calculated on different instrumentation set-up, namely UHPLC-ELSD and HPLC-ESI-MS, was mandatory to verify the proper functioning of the LRI library search, viz. the use of the LRI database, for achieving a reliable identification.

(120) Table 8 below contains the comparison for the most abundant TAGs identified in the three oils, along with the average and ΔLRI values, while FIG. 7 reports in a histogram the average values, each with the corresponding ΔLRI.

(121) TABLE-US-00008 TABLE 8 Average LRI, ΔLRI and LRI values calculated by UHPLC-ELSD and HPLC-ESI-MS. LRI TAG UHPLC-ELSD HPLC-ESI-MS LRI average ΔLRI LnLnLn 3712 3709 3711 2 LnLLn 3859 3853 3856 3 γLnLγLn 3902 3884 3893 9 LLLn 4018 4017 4018 1 LLγLn 4040 4025 4033 8 LnLnO 4030 4036 4033 3 LnPLn 4057 4046 4052 6 γLnγLnO 4092 4077 4085 8 γLnPγLn 4111 4091 4101 10 LLL 4179 4169 4174 5 LnLO 4202 4190 4196 6 LLO 4352 4343 4348 5 LLP 4373 4363 4368 5 OOLn 4375 4372 4374 2 OOγLn 4397 4377 4387 10 PγLnP 4443 4419 4431 12 OOL 4524 4514 4519 5 PLO 4555 4538 4547 9 SOLn 4584 4575 4580 5 SOγLn 4609 4594 4602 8 C22:1LL 4709 4704 4707 3 OOO 4732 4716 4724 8 POO 4764 4740 4752 12 POP 4782 4769 4776 7 SOO 4941 4940 4941 1 SOP 4959 4968 4964 5 C22:1OO 5102 5072 5087 15 C22:1OG 5239 5245 5242 3 C22:1OS 5267 5280 5274 7

(122) Conversely, the chromatographic method of the present invention can also be used to discriminate compounds having very similar or identical mass spectra. In summary, using retention indices in LC coupled with MS, particularly EI-MS or API-MS, does significantly increase the reliability of peak identifications by providing a second, independent experimental value.

OTHER EMBODIMENTS

(123) The embodiments and examples herein described have illustrative purpose only and they are not meant to limit the scope of the invention as claimed. The determination of other lipid species, such as free fatty acids, mono- and diacylglycerols, as well as sterols, carotenoids and waxes falls within the scope of the present invention. In fact, the inventors plan to make the method suitable for a holistic lipid profiling of different real-word samples in the near future. It is also obvious that the retention indices database of the present invention can be extended by including retention indices of other lipid species such as, for example, those of free fatty acids, mono- and diacylglycerols, as well as of sterols, carotenoids and waxes, for achieving a comprehensive database of all lipid species retention behaviour (e.g., t.sub.R and RI) of different real-world samples. The inventors are also planning to apply the retention index system to chromatographic method also suitable for the determination of polar lipids, such as phospholipids.

EXAMPLES

Samples Preparation

Example 1

Vegetable Samples Preparation

(124) Vegetable oil samples were prepared by dissolving 10 mg of oil in 10 mL of 2-propanol, except for borage oil for which 10 mg were dissolved in 1 mL of the same solvent.

Example 2

Milk Samples Preparation (by Folch)

(125) 10 mL of milk were treated with 40 mL of a chloroform: methanol mixture (1:2 v/v) and placed in an ice bath under magnetic stirring for 30 minutes; the content was transferred to a separating funnel and shaken vigorously for 5 minutes; after, the mixture was collected into different tubes and centrifuged at 3000 rpm for 15 minutes. Once obtained the separation of phases, the phases below (comprising the lipid fraction) were pooled together, the upper phases (aqueous) were added again to the separating funnel for a further extractions with 20 mL of chloroform: methanol mixture (2:1 v:v). The last step was replicated once again. The pooled organic phases were filtered on anhydrous sodium sulphate and evaporated using a rotary evaporator.

Example 3

Fish Samples Preparation (by Bligh&Dyer)

(126) 10 g of fish tissue were weighed and reduced to small pieces. The sample was extracted with 30 mL of a mixture chloroform: methanol (1:2 v:v) and homogenized by using an Ultra Turrax apparatus for 10 min. Then, 10 mL of chloroform and 18 mL of distilled water were added and re-homogenized for 1 min; the mixture was collected into different tubes and centrifuged at 3000 rpm for 15 min. The lower organic phases were pooled, while the aqueous upper phases were extracted again with 20 mL 10% (v/v) methanol in chloroform. Then, all the organic phases were pooled together, filtered on anhydrous evaporated using a rotary evaporator.

Instrumentation and Preferred Analytical Conditions

Example 4

HPLC-ESI-MS

(127) Analyses were carried out by using an Waters Alliance HPLC 2695 separation module consisting of a quaternary pump, autosampler and column thermostat, coupled to a Micromass Quattro micro API bench-top triple quadrupole mass spectrometer (Waters Associates Inc, Milford, Mass., USA). Separations were carried out on an Ascentis Express C18 100×2.1 mm (L×ID), 2.0 μm d.sub.p column (MilliporeSigma). Mobile phases were (A) acetonitrile: 10 mM aqueous ammonium formate 95:5 (v:v) and (B) 2-propanol under gradient conditions: 0-52.5 min, 0-55% B (held for 17.5 min). The flow rate was set at 400 μL/min with oven temperature of 35° C.; injection volume was 2 μL. MS acquisitions were performed using the Z-spray electrospray (ESI) source operating in positive ionization modes, under the following conditions: mass spectral range, 250-1250 m/z; event time, 1 s; desolvation gas (N2) flow, 700.0 L hr.sup.−1; no cone gas was applied; source temperature, 150° C.; desolvation temperature, 250° C.; capillary voltage, 3 KV; cone voltage, 80 V; extractor voltage, 3 V; RF lens, 0.2 V.

Example 5

UHPLC-ELSD

(128) Analyses were carried out by using a Nexera X2 system (Shimadzu, Kyoto, Japan), consisting of a CBM-20A controller, two LC-30AD dual-plunger parallel-flow pumps (120.0 MPa maximum pressure), a DGU-20A5R degasser, a CTO-20AC column oven, a SIL-30AC autosampler, and a SPD M30A PDA detector (1.8 μL detector flow cell volume). The UHPLC system was coupled to an ELSD (Evaporative Light Scattering Detector) detector (Shimadzu). Separations were carried out on two serially coupled Titan C18 100×2.1 mm (L×ID), 1.9 μm d.sub.p columns (MilliporeSigma, Bellefonte, Pa., USA). Mobile phases were (A) acetonitrile and (B) 2-propanol under gradient conditions: 0-105 min, 0-50% B (held for 20 min). The flow rate was set at 400 μL/min with oven temperature of 35° C.; injection volume was 5 μL. The following ELSD parameters were applied: evaporative temperature 60° C., nebulizing gas (N2) pressure 270 kPa, detector gain <1 mV; sampling frequency: 10 Hz.

Example 6

Identification of a Real-World Lipid Sample

(129) Different fish and vegetable oils were injected and positively identified using the chromatographic embodiment and the LRI method described in the present invention.

(130) FIG. 8 below shows the chromatogram of a sea bass lipid sample, extracted according to the procedure reported in Example 3 (Bligh&Dyer method), and analyzed according to the instrumental setup and analytical conditions reported in Example 5 (UHPLC-ELSD method). Peak identification based solely on the LRI identification method is reported in both the chromatogram and Table 9. A total of 30 TAGs were positively identified, being the difference between tabulated and experimental values ΔLRI≤15. Furthermore all identified compounds were confirmed by HPLC-ESI-MS analysis. The expansion of the database could ensure the identification of a major number of compounds.

(131) TABLE-US-00009 TABLE 9 PN, LRI.sub.tab, LRI.sub.exp and Δ values obtained for the identified peaks in a sea bass sample PN Compound name LRI.sub.tab LRI.sub.exp ΔLRI 38 LnLLn 3830 3821 −9 40 LLLn 3993 3980 −13 40 LnLnO 4011 3998 −13 40 LnPLn 4023 4017 −6 40 DhOPo 4135 4120 −15 40 DhPPo 4148 4141 −7 42 LLL 4160 4158 −2 42 LnLO 4192 4183 −9 42 LnLP + SLnLn 4217 4210 −7 42 DhOP 4246 4246 0 42 EpOO 4261 4264 3 44 LLO 4342 4330 −12 44 LLP + OOLn 4358 4345 −13 44 POLn 4383 4368 −15 44 DhSO 4399 4384 −15 44 OOPo 4485 4482 −3 46 OOL 4516 4505 −11 46 PLO 4539 4533 −6 46 SOLn 4563 4570 7 48 OLG 4703 4691 −12 48 OOO 4729 4716 −13 48 SLO 4746 4739 −7 48 POP 4776 4776 0 48 PPP 4784 4799 15 50 GOO 4905 4896 −9 50 SOO 4948 4937 −11 50 SOP 4961 4974 13 52 C22:1PO 5084 5099 15