Use of 13C-NMR as complementar identification tool in essential oil analysis

A program using C-NMR data was developed to improve GC/MS identifications of chemical constituents in essential oils. Compiled data uses 2290 and 2370 chemical shifts of 229 monoterpenes and 158 sesquiterpenes, respectively, stored into two reference libraries. Each carbon atom was codified using rules that correlates its chemical shift with structural features of the carbon hybridization and vicinity.


Introduction
Isolation of individual components from essential oils for NMR analysis, is somewhat difficult, time consuming and is not practical for complex oils. On the other hand, NMR essential oil analysis as a mixture is problematic because the chance of superimposing chemical shifts increases with the mixture complexity. Furthermore, many of the absorptions belonging to compounds present in low yields are not detectable due to background superposition. This explains why 13 C-NMR spectroscopy is largely used in identification and structural determination of pure compounds, but rarely applied when the analyte is a complex mixture like an essential oil [1][2][3]. It is also known that volatile oils constituents analysis only by mass spectroscopy, may present some problems [4], making desirable the use of other spectroscopic data to complement it. In this study is used GC/MS analysis of Alpinia speciosa essential oil presented in Table 3 with the help of an experimental procedure based on data of the 13 C-NMR overall spectrum of the oil presented in Fig. 2. The 13 C-NMR procedure compares each carbon chemical shift hybridization with the same parameters stored into libraries of mono and sesquiterpenes.
Prisco Bezerra, Universidade Federal do Ceará" under number 17624. Essential oil was first analysed using a HP-5971 GC/MS instrument equipped with a 25 m fused silica capillary SE-54 column (0.25 mm id), using helium as carrier gas (1 ml/min) and temperature programmed initially from 50 to 180 • C at 4 • C/min and then from 180 to 250 • C at 20 • C/min. 13 C-NMR spectrum of the oil was obtained on a Brucker A200 instrument (200 MHz, 12 hours of acquisition).

Reference libraries
Two references libraries were compiled using 2290 encoded carbon atoms from 229 monoterpenes and 2370 encoded carbon atoms from 158 sesquiterpenes.

13 C-NMR experimental data
In the Proton-Noise-Decoupled (PND) 13 C-NMR spectrum, recorded for Alpinia speciosa essential oil were detected 118 chemical shifts shown in Table 5. Chemical shifts hybridization patterns were deduced from Distortionless Enhancement by Proton Transfer (DEPT), at angles φ = 135 • and φ = 90 • .

Software
Computer programs to match experimental 13 C-NMR oil chemical shifts and carbon hybridization against those existing in the reference libraries, were written using FoxPro for Windows language.
All carbon atoms in the reference libraries were stored using a general code represented by C.αααα.rrr.ββ.γγ using letters separated by dots into five sub-codes as described below: The sub-code C identifies the substitution carbon atom pattern according to the rules shown in Table 1.
The sub-code αααα represents the four substituents directly bonded to the carbon atom C, fullfiled using the C and heteroatoms sub-codes defined in Table 1 completing the four digits with zero if necessary.
The sub-code rrr was created to designate rings that the carbon atom participates. The first two digits indicate the ring size in the increasing order of the carbon number. The last digit in this sub-code defines the stereochemistry of ring junction using capital letters Z for cis and E for trans, respectively.
The sub-code ββ represents the two most importants β effects induced by substituents linked to the carbon being analysed. When the hybridization of the selected C carbon is sp 2 , the two beta  substituents effects to be considered are two atoms apart from this carbon as shown in the example below [9]: The first digit in sub-code γγ represents the most important gamma substituent effect of carbon C and the second digit is reserved for the alpha (= a) or beta (= b) steric effect on this carbon. Another example for the sesquiterpene caryophyllene oxide illustrates how the codes were constructed and stored in the library (see Table 2).

Compound identification
The chemical composition of the A. speciosa essential was obtained by GC/MS library search best retrievals compatible with retention indices expressed as Kovats indices. Confirmations were made by visual comparison of each mass spectrum present in the oil with the corresponding mass spectrum in Refs [5][6][7][8] and by the 13 C-NMR procedure described here.

13 C-NMR matching program
13 C-NMR matching program applied to A. speciosa essential oil of were done when GC/MS identifications were already known. To illustrate how the program retrieves the library standards during comparisons, the chief constituent p-cymene and one of the minor constituents, caryophyllene oxide present in the oil were selected as example. As final result the program was able to detect all 13 C-NMR absorptions of p-cymene and 14 from the 15 chemical shifts of caryophyllene oxide as illustrated in Tables 3 and 4.

Application of 13 C-NMR matches to essential oil of A. speciosa
As stated before, compound identifications were first made by mass spectra library search by retrieving the best unknown structurally similar to the reference in the library, using GC/MS experimental data expressed in percent of agreement (0.0 = null match, 1.0 = full match), as shown in Table 5, column MIG.

13 C-NMR positive agreement
A positive 13 C-NMR agreement results when the chemical shifts compared for a pair of carbon atoms in the library and in the oil have exactly the same hybridisation pattern and their absolute chemical shift difference is no more than 1.5 ppm. A compound full agreement occurs when all  data being compared are retrieved from the library. In this case the comparison is expressed by the unity (1.00) otherwise it is expressed by the number of positive agreements divided by the number of chemical shifts existing in compound in the 13 C-NMR library.

13 C-NMR matches
The program make two runs using the positive agreement defined above. In the first run (match 1) 13 C-NMR library reference compounds corresponding to those identified by GC/MS, independently Table 6 Illustration how the 13 C-NMR program subtracts chemical shifts in the spectrum of the oil and identifies the ten major constituents PND HYB   Abbreviations: PND = proton noise decoupled; HYB = carbon atom hybridization; p-cym = para-cynene; 1,8-cin = 1,8-cineole; ter-4-ol = terpinen-4-ol; sab = sabinene; car-ox = trans-caryophyllene oxide; α-thu = α-thujene; lin = linalool; β-pin = β-pinene; α-ter-ol = α-terpineol; α-pin = α-pinene; ner = nerolidol; β-ter-ol = β-terpineol; unk = unknow; 9-ol = p-cymen-9-ol; myr = myrcene; shc = sabinene hidrate cis; sht = sabinene hydrate trans; β-car = beta caryophyllene; pip = piperitol; ac-b = bornyl acetate. of their yields in the oil, are matched against the overall spectrum of the oil. This inform how many absorptions of the compound being compared are present in the mixture. In the second run (match 2) comparisons follow decreasing order of the yields obtained by the GC/MS experiments shown in Table 5. When a positive agreement is found, the chemical shift just being compared is subtracted from the PND of the oil to avoid its use in a further comparison. Due to this approach, values in match 2 are represented by lower figures than those in match 1, but have the advantage of informing how many chemical shifts in the spectrum of the oil belong exclusively to the compound being matched (see Table 5, match 1 and match 2 columns). Table 6 shows in the PND column, the 118 chemical shifts found in the 13 C-NMR spectrum of the oil. When a chemical shift is present in all columns mean that this absorption does not belong to any of the ten major compounds listed on the head of Table 6. An asterisk in a given column means that the corresponding chemical shift belongs to the compound in the head of the column. An asterisk is always followed by minus (-) signal to indicate that chemical shift was subtracted from the PND, to avoid their use en another comparison. The last column in Table 6 shows abbreviations of all compounds identified in the oil as well as those not identified represented by the abbreviation unk.

Conclusions
From Table 5, considering this particular essential oil analysis, some conclusions may be anticipated: 1. Compounds with high GC/MS MIG values are the best candidates for confirmation using match 1 and match 2.
3. Figures with high MIG value, present in the oil in low yieldα-thujene, α-pinene, β-pinene, myrcene, linalool, α-terpineol -have their identifications confirmed by high value of match 1 (>0.70). 4. Even low values in match 2, but with significative figures in match 1 for compounds in low yield, are of great importance for identification purposes because identify in the mixture data belonging exclusively to the compound being compared.