Controlling Factors Determining the Selective HSCN Addition to Double Bonds and Their Application to the Synthesis of 7-Isothiocyano-7,8-a-Dihydro-Bisabolene.

The reactivity of terminal and trisubstituted double bonds of monoterpenes with HSCN has been examined by GC giving evidence that kinetics is responsible for the chemoselective addition to terminal double bonds in terpenes. The results show that the addition to the terminal double bond is about 17 times faster than for trisubstituted double bonds and that the presence of the first SCN group in the molecule prevents a second addition. The presence of a hydroxyl or methoxy group in the molecule, decreases the reaction kinetics. Based on these kinetic experiments a two steps synthesis of the natural product 7-isothiocyano-7,8-dihydro-a-bisabolene using bisabolol as starting material, was planned and successfully accomplished.


Introduction
Among the sulfur containing natural products the biologically active isothiocyanosesquiterpenes isolated from sponges (order Halicondrida, Axinellida and Littristida) 1,2 have attracted the attention of several synthetic chemists.A critical analyses of some isothiocyanosesquiterpene total syntheses 3 revealed that a good method to selectively introduce the NSC group at a quaternary position was lacking.The addition of the HSCN to double bonds, was successfully applied in our group, to obtain mono and sesquiisothiocyanoterpenes proving to be a straightforward methodology showing regio and chemoselectivity.In some instances stereoselectivity was also observed.From the initial experiments, it was inferred that the observed chemoselectivity could be explained in terms of kinetics but no specific experiments were performed.
Unraveling the chemoselectivity of the HSCN addition to terpenes is the objetive of this present paper .

Results and Discussion
Limonene (1) and dihydrolimonene (2) were proposed as starting material for the kinetic study of the HSCN addition.These two monoterpenes were considered to be ideal to provide the answers to two basic questions: a) is HSCN addition to a terminal double bond indeed faster?and b) are any double addition products formed in detectable amounts?
The answer to question a is obtained from kinetic experiments.Thus standards of the 3/4 (thermodynamic products) and 3a /4a (kinetic products) were obtained using the previously described methodology 4,5 with a large excess of in situ generated HSCN (Figure 1).All standards were carefully characterized by spectroscopy.Gas chromatographic monitoring of HSCN addition to an equimolar mixture of 1 and 2 provided a data ensemble (Table 1) that clearly differentiated between a fast terminal and a slow trisubstituted double bond reaction.
Quantitative treatment of these data could be performed by applying the Sharpless equation 6 assuming a pseudo first order reaction (HSCN present in large excess).
However, this quantitative method requires accurate evaluation of the conversion extent.This particular information could not be accessed due to serious allergic reactions of the chemists directly and indirectly working in this project while isolating compounds 3, 3a, 4 and 4a.On the other hand, Rakels' equation 7 (Equation 1) provides an excellent means to evaluate the relative quantities of k 1 and k 2 and conversion data are not required.This equation was adapted to our experiments taking into consideration that when the GC detector response is corrected for the substrates and for the products they will behave as an enantiomeric pair.Thus using data obtained after 65 min of reaction,   Thus the chemoselectivity of HSCN addition to the terminal double bond of 1 and other terpenes can be partially explained in terms of kinetics 5 .
The answer to question b came from the fact that no additional products were detected in the reaction monitored by GC and GC/MS.A second factor has to be invoked for the non formation of the double addition products.The literature 8 has previously reported that in HSCN addition to double bonds, the presence of polar groups play a deactivating role.By monitoring the addition reaction of HSCN to an equimolar mixture of 2 and 5 during a reaction time of 24 hours, dihydrolimonene (2) was totally consumed while no addition product from terpineol (5) was detected (Figure 2).Consequently the factors determining the chemoselectivity are: a) faster addition to the terminal double bond and b) interference of the first SCN group in the molecule, preventing a second addition.
In order to assess the scope of the polar group interference, addition of HSCN to (±)-O-methylcadinol (6), cadinol 9 (7) (equatorial OH), cadinol 9 (10) (axial OH), isothiocyanocadinene (13), thiocyanocadinene ( 14) and (-)-bisabolol 10 (8) were carried out.Compounds 6, 7, 13 and 14 did not produce addition products in detectable amounts while (-)-8 produced 9 and 9a and 10 produced 11 and 11a in 1:1 ratio (Figures 2 and 3).Therefore, polar groups can a priori be divided into two groups: a) polar groups (e.g.OH) capable of hydrogen 'bonding and located in the vicinity of the double bond which, though decreasing the overall reaction kinetics, can favour one addition among other alternatives.The selective addition to only one of the two trisubstituted double bonds of 8 is a good example.Another example is cadinol (10) (Figure 3) which produced 11 and 11a.b) polar groups which decrease the rate of the addition and provide no counterbalancing effect, so that most of the HSCN is polymerised before any addition occurs (e.g.-OMe group and an OH group with no appropriate stereochemistry to favour one addition, compounds 6 and 7).The interference of the polar groups was well documented with compounds 6, 7, 8 and 10 but we felt that the interference of the first SCN ought to be better investigated.Thus addition to cadinene (12) (Figure 3) was monitored by GC and GC/MS.As expected, addition occurred at the terminal double bond.The reaction solution was then filtered and all the polymerised HSCN was separated.Freshly prepared HSCN was then added to the filtered reaction solution.GC monitoring revealed that the 13 plus 14 mixture was transformed into the thermodynamic products 13a and 13b but no double addition occurred.Thus the SCN and NCS groups can be classified as polar substituents belonging to group b.Aiming at the application of these newly acquired insights about HSCN addition to terpenes, the synthesis of 7-isothiocyano-7,8-dihydro-a-bisabolene (16), isolated from Halichondria sp. 11, in two steps from (-) -bisabolol (8), was proposed as a challenging synthetic target (Figure 4).It should be mentioned that the natural product is (+)-1R,7S and the use of (-)-(1S,7S)-bisabolol (8) would yield the enantiomeric series.Nevertheless, (-)-bisabolol (8) dehydration 12 provided a mixture of 15, 15a in a 1.5:1 ratio based on GC and 1 H NMR data, [d 4.73 (H-14 of 15), 5.00 (H-8 (15a) and H-10 (15 and 15a)]. 13o attempts were made to separate these structurally related isomers since according to our results only compound 15, possessing the terminal double bond would react rapidly and the remaining hydrocarbons would be easily separated by chromatography after the reaction.This indeed occurred and the addition products 16 and 16a were obtained in 47% yield.If one considers that 15 was only 60% of the starting material (obtained by GC analysis) the calculated yield is 77%.Comparison of the spectroscopic data of the synthetic 16a and of the natural product showed equal proton and C-13 chemical shifts which is expected for enantiomeric pairs thus confirming the synthetic success.
Finally it can be concluded that Eq. 1 allows relative quantitative evaluation of a reaction kinetics by GC/FID after calibration response of the reagents and the products.There is no need to carry the reaction to completion.
Application of Eq. 1 to our data revealed that HSCN addition to terminal double bonds is faster and responsible for the chemoselectivity.The polar group interference in the HSCN addition to terpenes was assessed through four additional reactions and the synthesis of 7-isothiocyano-7,8-dihydro-a-bisabolene in two steps from bisabolol was the final proof of the chemoselectivity of this reaction.

Experimental
FT-IR Spectra were recorded with a Perkin Elmer 298 spectrophotometer as film on KBr cells. 1 H NMR spectra were recorded with Varian GEMINI 300 (300.1MHz,Varian) or Bruker AC 300P (300.1 MHz) spectrometers CDCl 3 was used as the solvent, with Me 4 Si (TMS) as internal standard. 13C NMR spectra were obtained with a Varian GEMINI 300 (75.5MHz) or a Bruker AC300P (75.5MHz) spectrometers.CDCl 3 (77.0ppm) was used as internal standard.Methyl, methylene, methyne and carbon non bonded to hydrogen were discriminated using DEPT-135° and DEPT 90° spectra (Distortionless Enhancement by Polarization Transfer).2D NMR spectroscopy was performed with standard H,H correlation and H,X correlation pulse sequences available in the spectrometers.Optical rotation values were measured with a Polamat A polarimeter and the reported data refer to the Na-line value using a 1 dm cuvette.The GC/MS analyses were carried out using a HP-5890/5970 system equipped with a J&W Scientific DB-5 fused silica capillary column (25 m x 0. The numbering systems adopted to assign protons and carbon signals in the NMR spectra is depicted in Figures 1, 2 and 3 which in some structures is different from the numbering following IUPAC nomenclature.The IUPAC names are in brackets in the experimental section. Thiocyanic Acid: In an Aldrich atmosbag, a slurry of powdered KSCN(7.3g, 75.0 mmol) in 30 cm 3 of CHCl 3 was triturated with 11.2g (82.0 mmol) of KHSO 4 in a mortar for 5 min.The HSCN chloroform solution was decanted and an additional 10 cm 3 of CHCl 3 was added to the solid mixture and then decanted.The combined solutions totalled 30 cm 3 .

Figure 2 .
Figure 2. Terpenes possessing polar groups that did not undergo HSCN addition.