Geochemical Transactions, 2001, 3 DOI: 10.1039/b102650j	Paper
The nature and fate of natural resins in the geosphere Part XI.† Ruthenium tetroxide oxidation of a mature Class Ib amber polymer

Ken B. Anderson

Received 22nd March 2001 , Accepted 4th June 2001

Published on the web 28th June 2001

The results of ruthenium tetroxide (RuO₄) oxidation of a mature Class Ib amber polymer are reported and discussed. These data indicate that the residual double bond present in mature Class I ambers is not located in the A/B ring structure of these materials and that C17 of the original labdanoid precursors is retained in mature Class I ambers as a methyl group. These data also suggest that the reaction which results in formation of the residual unsaturated structure in mature ambers also results in a second covalent connection between the A/B ring system and the polymer backbone, probably through C8 of the original labdanoid structure.

Introduction

Fossil resins (ambers), are common sedimentary constituents in many areas, and occur in great abundance in some deposits, (e.g., around the Baltic region, where amber has been commercially produced for centuries and in the Dominican Republic where commercial deposits are also mined). The majority of significant deposits are Tertiary but older deposits, stretching back to at least the Triassic, are known.

Fossil resins, (ambers) are common sedimentary constituents in many areas, and occur in great abundance in some deposits, (e.g., around the Baltic region, where amber has been commercially produced for centuries and in the Dominican Republic where commercial deposits are also mined). The majority of significant deposits are Tertiary, but older deposits stretching back to at least the Triassic are known.

Ambers are classified chemically on the basis of their molecular structure.^2,3,4 Class I ambers are derived from higher plant resins based primarily on polymers of labdatriene diterpenes. When the plant exudes resin, (usually in response to injury or stress of some form), these compounds undergo polymerization across the terminal double bond located in the side chain to give a general 14,15-poly(labdatriene) polymer structure⁵ (numbering used is the accepted numbering of labdanoid diterpenes). In many cases, a variety of related labdanoid precursors are incorporated into the developing polymeric structure resulting in a final copolymeric structure incorporating labdanoid carboxylic acids, alcohols and hydrocarbons. Ambers based on both regular and enantio labdanoids are well known in the geosphere,⁶ and sub-classification of Class I ambers is based on this distinction (and on the incorporation, or lack thereof, of succininc acid).²

To further elucidate the details of this process, a mature Class Ib amber (Upper Cretaceous, Yantardakh Hill region, Siberia) has been subjected to ruthenium tetroxide (RuO₄) oxidation. This reagent aggressively cleaves carbon–carbon double bonds but does not attack saturated structures. In order to simplify interpretation of the resulting data, the sample was first extracted to isolate a soluble polymer phase¹¹ which was as free as possible of extraneous occluded materials. The results of these investigations are reported herein.

Experimental

Extraction

Amber (251.9 mg) was placed in a micro Gregar extractor and extracted with 1 1 (v/v) CH₂Cl₂ n-pentane to remove occluded materials. After 24 h, extraction was stopped and the CH₂Cl₂–n-pentane solution removed and replaced with CH₂Cl₂ CH₃OH (1 1 v/v) and the extraction recommenced. After a further 24 h, the CH₂Cl₂–CH₃OH soluble phase was rotary evaporated to near dryness, then taken up to the extent possible in 10 mL CH₃OH and fully precipitated by slow addition of cold H₂O. The resulting precipitate was recovered by filtration and washed with cold H₂O. Residual insoluble materials were also collected, and washed with CH₃OH–H₂O. Both products were then dried overnight in vacuo at 60 °C to give 149.6 mg soluble polymer (60% yield relative to original amber) and 15.2 mg insoluble residue (6% yield relative to original amber).

RuO₄ oxidation

Soluble polymer (24.8 mg) was placed together with NaIO ₄ (220 mg) and RuO₂·xH₂O (9.6 mg) in a 25 mL round-bottom flask to which CCl₄ (1.0 mL), CH₃CN (5 mL) and H₂O (4 mL) were then added. The resulting mixture rapidly yellowed and was allowed to stir for 4 d, after which time the reaction was quenched by addition of CH₃OH (0.5 mL). The entire crude product was then evaporated to dryness, then taken up in CH₃OH (4 × 10 mL washes) and filtered to remove insoluble byproducts (largely RuO₂, NaIO₃ and any unreacted NaIO₄). To this MgSO₄ was added to remove residual H₂O and the resulting mixture refiltered to give a clear CH₃OH solution which was then evaporated to ca. 20 mL. To this solution 2 mL of BF₃–CH₃OH complex was then added and the reaction was refluxed for 2 h, then allowed to stand overnight, after which 5 mL saturated aqueous NaHCO₃ was added to terminate the reaction. This product was then taken up in diethyl ether–CH₃OH (1 3 v/v 40 mL) filtered over MgSO₄, evaporated to dryness, washed again with H₂O, then finally recovered by filtration and dried at 60 °C overnight. Yield ca. 70% (by mass).

¹H and ¹³C NMR spectroscopy

NMR experiments were performed on a Bruker model DMX 500 NMR spectrometer (11.7 T). With the use of the nitrogen pre-cooler, heater coil, and the variable temperature controller, the temperature was stable at 294.3 × 0.1 K. ¹³C NMR spectra were recorded with 22,000 acquisitions, 10 s recycle delay, 9 ms pulse width (70°) and 33 kHz sweep width using a two channel 10 mm broadband, direct detection, variable temperature probe with proton decoupling and ²H lock at 76.773 MHz.¹H experiments were carried out using a three channel 5 mm inverse detection, three axis gradient, variable temperature probe with ²H lock at 76.773 MHz. ¹H spectra were recorded with 64 acquisitions, 4 s recycle delay, 4 ms pulse width (70°) and 6 kHz sweep width. Solvent used for both ¹³C and ¹H experiments was CD₂Cl₂–CD₃OD ca. 1 1 vol.

Pyrolysis GC-MS and GC-MS analyses

Both the isolated soluble polymer and the product of its RuO ₄ oxidation (which was found to be unamenable to analysis by gas chromatography—no significant products observed) were further analyzed by Py-GC-MS. Details of the analytical system used have been reported elsewhere.¹² Briefly, T_pyrol = 480 °C (10 s). Chromatographic conditions: 60 m DB-XLB (0.25 mm id, 0.25 m phase thickness), T_init = 40 °C (4 min), ramp at 4° min¹, T_final = 310 °C (16 min). For analysis of the isolated polymer, TMAH was added as described elsewhere.¹² No TMAH was used in Py-GC-MS analyses of the RuO₄ oxidation product of the soluble polymer.
GC-MS analyses were conducted using an HP 5972 MSD coupled with an HP 5890 (II) equipped with an automated direct on-column injector. GC-MS analyses reported herein (Data set 1) were carried out using a 60 m DB-5 MS column (0.25 mm id, 0.25 m phase thickness) programmed as follows: T_init = 140 °C (4 min), ramp at 4° min¹, T_final = 300 °C (10 min).

Results and Discussion

¹³C NMR analysis of the soluble polymer prior to oxidation indicates terminal CH₂total other CC (excluding aromatic C) of approximately 1 7.5, indicating that maturation related transformation of the initial exomethylene double bond is advanced to essentially complete in this sample. (Click here to access these data—Chime plug-in required). ¹H NMR spectra for this sample are also consistent with this conclusion. (Click here to access these data—Chime plug-in required).

Results of Py-GC-MS analyses of the isolated soluble polymer of the amber used in this study are given in IDS-2.

All of the products observed in these data belong to one of the series of regular bicyclic compounds derived from the A/B ring system of various regular labdanoid monomers. These compounds are characteristic of the pyrolysates of Class Ia and Ib ambers.^2,6 Structures in which C11 is retained predominate over those in which this carbon is absent, as is typical for mature Class I ambers.² These data are, not surprisingly, closely comparable with data from analyses of the whole amber from this site,³ the only significant difference being the presence in significant abundance of bicyclic methyl ethers, which are undoubtedly an artefact of the isolation procedure used. These data also show that this product is free of occluded materials, especially abietane-type diterpenes, which are observed in the pyrolysate of the whole amber³ and which may have complicated interpretation of subsequent data from RuO₄ oxidation studies.

Data from Py-GC-MS analysis of the RuO₄ oxidation product of the amber polymer described above, are given in IDS-3.

Although not all of the observed products can be unambiguously assigned at this time, data from the hydrolysis–methylation of sclareolide (Data set 1) and comparison of the data given in Data set 3 with literature data¹⁴ do allow the majority of products, including most of the major products, to be assigned with a high degree of confidence. All of these compounds (Data set 3, 1–3 and 6–8) belong to two homologous series of compounds which are clearly derived from, and which preserve the A/B ring of the original labdanoids.

The structural characteristics of these compounds are very informative in a number of ways. Firstly, as just noted, these compounds all preserve intact the original labdanoid A/B ring structure. Given the high yield recovered from the oxidation procedure, this indicates that the residual double bond present in this amber is not located in this ring system, since had it been so, these rings would have been cleaved by the oxidation procedure. It can also be concluded that the final double bond is not directly attached to this ring system since if it were then one would anticipate keto products at the point of attachment, and no such products have been identified.

It is also informative that all of these products are themselves unsaturated. Clearly this is a result of the pyrolytic analyses used in this investigation since olefinic structures are very susceptible to cleavage by RuO₄ and NMR analyses of this product prior to pyrolysis confirm the absence of CC unsaturation. All of these products are unsaturated around C8. This suggests that prior to pyrolysis (and after oxidation) this carbon is likely to be functionalized and that on pyrolysis this functionality is being lost, probably through dehydration or decarboxylation or a similar mechanism.

In all of these products C17 of the original labdanoid is retained. In the original labdanoid structure this carbon is present in an exomethylene structure. Progressive loss of exomethylene character is one of the defining characteristics of maturation of these polymers,^2,7,8,9 and it might be (and in fact has been^8,9,15 ) postulated that this reflects cross-linking or condensation reactions directly involving this structure. However, the preservation of this carbon in all cases as a methyl group argues against this possibility since such reactions would retain this carbon as a methylene structure and it is difficult to postulate any mechanism by which such a structure can be universally converted to a methyl group upon pyrolysis. It is also noteworthy that all of the characteristic bicyclic structures observed in the pyrolysates of Class I ambers also retain this carbon as a methyl group, which supports the view that this carbon is retained as a methyl group in mature Class I ambers.

If it is correct that C17 is retained in mature Class I ambers as a methyl group, then this implies that isomerization of the initial exomethylene structure is an early step in the maturation of these materials. However, as noted above, the data given in Data set 3 preclude the presence of a double bond in, or on the labdanoid A/B ring system in mature samples. How then, can these observations be reconciled? One additional observation is helpful in this regard—the product of RuO₄ oxidation of this polymer was not amenable to analysis by GC-MS. Although it was not possible to determine absolute molecular weights for the RuO₄ oxidation products described herein, this observation suggests that this product is still a high molecular weight material.

If maturation of polymers with an initial 14,15-poly(labdatriene) structure results in a structure with one double bond, which is not located in nor directly attached to the A/B ring systems of the labdanoid monomers, then it follows ex necessitate rei, that the double bond must be located somewhere in the side chain structure. However, one would expect oxidation of such a structure to give volatile, low molecular weight products due to cleave of the linkage between the polymer backbone and the A/B ring system, and this is observed on oxidation of immature samples (Anderson, unpublished results). The absence of volatile products from the RuO₄ oxidation of this mature sample therefore suggests that covalent connection still exists between the polymer backbone and the A/B ring system after oxidation.

Hence, we are left with the following:
(i) Maturation of Class I ambers results in a structure with 1 double bond per original monomer.^8,9
(ii) This residual double bond is not structurally equivalent to either of the initial double bonds,¹⁰ nor is it located in, or directly attached to the labdanoid A/B ring system of the original monomer.
(iii) C17 of the original labdanoid structure is retained in the mature structure as a methyl group.
(iv) In addition to the connection through the original side chain, maturation results in a new covalent connection between the polymer backbone and the labdanoid A/B ring system.

Acknowledgements

The support of the US Department of Energy, Office of Science. Basic Energy Sciences Division of Chemical Sciences, Geosciences, and Biosciences, under contract number W-31-109-ENG-38 is gratefully acknowledged. Dr. John V. Muntean (Chemistry Division, Argonne National Laboratory) is thanked for his assistance in obtaining the NMR data described in this report. Chemical structures are rendered in this report using the MarvinView() Java Applet developed by Peter Csizmadia and coworkers at ChemAxon Ltd. Systematic names for structures reported herein are generated using AutoNom2000 (MDL Inc.) using CAS naming conventions. The comments of two anonymous reviewers are also gratefully acknowledged.

† For Part X of this series see ref. 1.