Norbornadiene as a Universal Substrate for Organic and Petrochemical Synthesis

A wide range of rare polycyclic hydrocarbons can be obtained through catalytic processes involving norbornadiene (NBD). The problem of selectivity is crucial for such reactions. The feasibility of controlling selectivity and reaction rate has been shown for cyclic dimerization, co-dimerization, isomerization and allylation of NBD. Kinetic rules have been scrutinized. Consistent mechanisms have been proposed. Factors affecting directions of the reactions and allowing us to obtain individual stereoisomers quantitatively, have been established. A series of novel unsaturated compounds has been synthesized; they incorporate a set of double bonds with different reactivity and can find an extremely wide range of applications. Eurasian ChemTech Journal 3 (2001) 73-90  2001 al-Farabi Kazakh St.Nat. University Introduction The importance of NBD and its derivatives in various fields of human activity is growing. More and more new applications are being discovered for these compounds. Since they were first obtained less than 50 years ago, these substances have been successfully utilized in medicine, agriculture, rocketry, syntheses of polymers with unique properties, microelectronics, and as solar energy converters. The number of patents concerned with synthesis and application of NBD derivatives and norbornene-2 (NBN) had exceeded 100 hundreds by the year 2000. Due to their unique structure, compounds with NBD and NBNmoieties are achieving the leading positions in contemporary chemistry and chemical technology [1-3]. It should be strongly emphasized that NBD itself and some of its simplest derivatives are obtained from chemicals, produced on a large scale during oil processing [4]. They are cyclopenta-1,3-diene (CPD), acetylene, alkenes, and alkadienes with various structures. Production of CPD can easily be combined with synthesis of other petrochemical products, for example ethylene. Today, a large proportion of CPD cannot be duly utilized, so it is very important to search for new ways of its utilization. Even though NBD has extremely rich synthetic possibilities, its utilization as a universal substrate is rather limited, since NBD-derivatives can form all kinds of isomers: skeleton, regio, stereo, enantio ones. A resultant mixture of isomeric products is often difficult to separate and analyze, and the consumption of reagents may become excessive. All this restricts widespread use of NBD. Metal-complex catalysis is very interesting and attractive for solving various problems connected with selectivity [5]. By applying its tenets and methods to reactions involving NBD and its derivatives, by scrutinizing mechanisms of reactions, we can control structure and range of regioand stereo isomers, synthesize new compounds of this type, and can make their production sensible from technological and economical points of view. The present article bases on several research directions concerned with application of the above-mentioned methods in some of the most promising processes involving NBD.


Introduction
The importance of NBD and its derivatives in various fields of human activity is growing. More and more new applications are being discovered for these compounds. Since they were first obtained less than 50 years ago, these substances have been successfully utilized in medicine, agriculture, rocketry, syntheses of polymers with unique properties, microelectronics, and as solar energy converters. The number of patents concerned with synthesis and application of NBD derivatives and norbornene-2 (NBN) had exceeded 100 hundreds by the year 2000. Due to their unique structure, compounds with NBD and NBNmoieties are achieving the leading positions in contemporary chemistry and chemical technology [1][2][3].
It should be strongly emphasized that NBD itself and some of its simplest derivatives are obtained from chemicals, produced on a large scale during oil processing [4]. They are cyclopenta-1,3-diene (CPD), acetylene, alkenes, and alkadienes with various structures. Production of CPD can easily be combined with synthesis of other petrochemical products, for example ethylene. Today, a large proportion of CPD cannot be duly utilized, so it is very important to search for new ways of its utilization.
Even though NBD has extremely rich synthetic possibilities, its utilization as a universal substrate is rather limited, since NBD-derivatives can form all kinds of isomers: skeleton, regio, stereo, enantio ones. A resultant mixture of isomeric products is often difficult to separate and analyze, and the consumption of reagents may become excessive. All this restricts widespread use of NBD. Metal-complex catalysis is very interesting and attractive for solving various problems connected with selectivity [5]. By applying its tenets and methods to reactions involving NBD and its derivatives, by scrutinizing mechanisms of reactions, we can control structure and range of regio-and stereo isomers, synthesize new compounds of this type, and can make their production sensible from technological and economical points of view.
The present article bases on several research directions concerned with application of the above-mentioned methods in some of the most promising processes involving NBD. lytic methods of their synthesis under homogeneous conditions, they take high-vacuum equipment and usage of purified inert gases. All cited compounds were identified by means of: chromato-mass spectrometry with use of chemical ionization and electron blow as well («Kratos MS-80» and «HP-5989B MS MG»); 1 H, 13 C and 31 P NMR spectroscopy («Bruker WP-250», «Bruker DPX-300», «Bruker MSL-200»); infra-red Fourier-transform spectroscopy («Bruker IFS-113V» supplied with chromatographic device «Karlo-Elba Strumentation 4200»); gas-liquid chromatography («Khrom-5» with a range of capillary and preparative columns). Structure of catalysts was investigated by X-ray photoelectron spectroscopy («Riber LAS-4000») and electron spin resonance spectroscopy («Bruker ER-200»).

Valence isomerization of NBD
Valence isomerization of NBD into quadricyclane (Q) has a considerable practical importance [6]. This reaction draws attention, since it allows us to accumulate solar energy through a cyclic process (reaction 1).
The characteristics of phototransformaiton of NBD into Q can be substantially improved through introduction of substituents into the NBD molecule. For instance, when both an electron donating and an electron withdrawing moieties (Ñ 6 Í 5 and ÑÎÑ 6 Í 5 respectively) are simultaneously introduced into one of the double bonds, the limit of adsorption can be moved into the long-wave area up to 350 nm or, if the aryl substituent gets additionally modified, up to 400 nm, with the quantum yield increasing up to 0.3÷0.6 [7].
Further introduction of substituents into different double bonds allows us to vary λ from 350 to 450, with a quantum yield being up to 0.96; the best results were attained when electron-acceptor substituents were introduced into one of the double bonds, and electron-donor ones into the other.
Compounds of the NBD-series with a similar structure are suitable for practical application, even though they have shortcomings. The most serious of them are the decrease in the quantity of accumulated energy per 1 g of compound (due to the increase in the molecular weight of substituted NBD) and complexity of their synthesis.
One more general drawback is side reactions with NBD. Even though the yield of side products is negligible from the conventional point of view (hundredths of one per cent per one working cycle), they accumulate as the cycle NBD↔Q repeats over and over again. This substantially deteriorates the performance of the system.
To increase the number of working cycles, NBD molecules can be covalently bound to a polymeric matrix. For instance, application of polymethylacrylate films increases the number of working cycles up to 10 3 -10 4 , with λ and Ô remaining high [8].
Phototransformation of unsubstituted NBD into Q features a low quantum yield. This can be considerably increased through applying sensitizers. The best results were obtained for Cu (I) salts (Φ=0.2) and phenylketones (acetophenone, benzophenone Φ=0.5÷0.9). However, even these systems have imperfections. First, they work only in the UV spectrum area. Second, Cu (I) complexes turn into Cu (II) ones, which show no photoactivity. Third, ketones chemically react with NBD to give photo-adducts. All this prevent such stabilizers from practical application. These obstacles may probably be overcome through use of natural or artificial polyheterocyclic sensitizers such as porphyrins [9].
A system for solar energy accumulation can only In this photoreaction, solar energy is accumulated due to the formation of a methastable structure incorporating highly-strained moieties: one cyclobutane and two cyclopropane rings, which results in an unusually high thermal effect of the reverse dark reaction (110 kJ/mol). This system is also characterized by the following positive qualities: (i) the direct photochemical reaction can be sensitized; (ii) the reverse reaction has a high activation barrier (its half-life period is 14 hours at 140°C), but can be accelerated catalytically; (iii) NBD and Q are liquids, which is convenient from the technological point of view.
The process has some shortcomings, too: NBD adsorbs in the short-wave area (up to 300 nm) and the quantum yield of the direct photoreaction in the hν NBD Q catalyst (1) be created if the product of the photoreaction is chemically stable. The high energy barrier for the thermal transition NBD→Q results from orbital symmetry rule for [2σ+2σ]-addition. This barrier can be made lower through using compounds of transition metal as catalysts.
Of these substances, flat-square cobalt (II) complexes are the most promising. They are more active than compounds of other metals under both heterogeneous and homogenous conditions. Besides, no side reactions occur when Co (II) compounds are used. However, under isomerization conditions, many of Co (II) complexes oxidize easily into Co (III), which in turn leads to formation of NBD dimers and trimers. From this point of view, Co(II) porphyrin complexes are considered the best nowadays. When covalently bonded to a polymeric support or applied to inorganic carriers (activated carbon, alumina), they can easily be regenerated and are suitable for practical use.
The global ecological problem imposes stricter requirements on chemical production processes. Hence photochemical methods can play an important role, since they are able to provide a means to absorb solar energy and utilize it in chemical transformations. Light is a kind of an inertialess chemical reagent creating no wastes. However, photochemical processes only have minor importance now, chiefly because some complicated technical problems have not been solved yet.
All the above-described is completely correct for the NBD-Q system. Its practical significance is obvious. NBD-based pilot power units have already been being developed in several countries.
However, economical obstacles hinder the widespread use of heat energy emitted through the chemical transformation of Q into NBD. At the moment heat (as water steam) generated though this process is 50 to 100 times more expensive than that obtained through conventional methods. The mentioned systems need further improvements: the number of working cycles should be increased up to 10 4 and higher, the quantum yield and conversion of NBD per cycle should be increased, and the cost of synthesis of NBD with required spectral characteristics must be reduced. However, the creation of portable units may be sensible even now: they can be used in sunny regions located far away from other energy sources, or for space satellites.
Dimers of group IV (15)(16)(17) obtained in the presence of [Rh(PPh 3 ) 3 Cl] are probably derivatives of 9. We can suppose that they appear due to incorporation of the Rh(I) complex in the cyclopropane ring of the nortricyclyl moiety 9 with a subsequent transfer of a hydrogen atom from the bridge carbon.
Catalysts are necessary for NBD cyclodimerization to take place. They are usually complexes of Ni, Co, Fe and Rh in lower oxidation states. Some examples of utilization of Cr, Ti, Pd and Ir compounds were reported. The type of the central atom substantially influence the direction of cyclodimerization of NBD, i.e. the structural selectivity of the process.
Iron-containing catalytic systems can act in a number of different ways. Various kinds of cycloaddition can be realized through varying ligands, reducing agents, and reaction conditions. For example, iron carbonyl complexes Fe(CO) 5 , Fe 2 (CO) 9 , Fe 3 (CO) 12 catalyze all routes of NBD cyclodimerization and give rise to a mixture of products, chiefly isomers 1, 8 and 9.
If several nitrosyl ligands are substituted for carbonyls, the activity and selectivity of ferruginous catalysts can be enhanced. For instance, when Fe(CO) 2 (NO) 2 is used, dimerization only proceeds through the [2π+2π]-direction to give dimers 1 and 3. The unsaturated complex Fe(NO) 2 formed in situ via a reaction of nitrosylferrates with NBD or its dimers, is probable the key intermediate through which NBD dimers are formed.
That the moiety Fe(NO) 2 is catalytically active is indirectly confirmed by a series of electrochemical experiments where a range of nitrosyl iron salts were reduced. Catalysts generated this way yield dimers 1 and 3 only, independent of the structure and type of the initial complex.
The type of reducing agent almost does not affect the yield of dimers and the selectivity of the process. However, when powdered zinc is used, the conversion of NBD depends on the type of solvent. Acetone and THF are more favorable for the reaction than toluene, methanol or acetonitrile [10].
It is interesting, that the character of cycloaddition can be changed completely through adding an equimolar quantity of BF 3 ·OEt 2 : in CH 2 Cl 2 , the only product is [4π+4π]-isomer 11 (Binor-S). When twocomponent systems like Fe(acac) 3 -AlEt 3 are used, a mixture of dimers 1, 3, 7 è 9 is usually formed [12]. The content of isomers depends on the solvent type and how the catalyst was synthesized [13][14].
The yield of hexacyclic dimers can be increased through adding phosphine ligands to this catalytic system, and substituting reducer AlEt 2 Cl for AlEt 3 . When a chelate bis-(diphenylphosphino)ethane (diphos) is used, the yield of 9 increases up to 80-90%. In the presence of triphenylphosphine the major product is dimer 7 [15].
These systems generally manifest a high activity, yield and turnover. Their shortcomings are low selectivity to individual isomers and non-repeatability of results.
The selectivity of nitrosyl Co complexes is lower than that of analogous ferruginous catalysts. Even though Ño(CO) 3 NO is isoelectronic with Fe(CO) 2 (NO) 2 , NBD cyclodimerization in its presence gives not only 1, but also a mixture of hexacyclic hydrocarbons. Isomer 1, however, can be synthesized with a selectivity up to 98% when complexes Co(NO) 2 Cl or Co(NO) 2 Br promoted with AgBF 4 or NaBPh 4 are used [16].
A series of catalysts like M[Co(CO) 4 ] n (M = Zn, Cd, Hg) can stereospecifically dimerize NBD into Binor-S (11) with a yield of 95% [17]. These compounds only manifest their catalytic activity in the presence of Lewis acids (BF 3 , SbF 3 , AlBr 3 , BF 3 ·OR 2 ) as co-catalysts with a ratio of 1 to 8. Lewis bases (pyridine, triethylamine) favor formation penta-and hexacyclic dimers [17]. Acetylacetonate Co systems reduced by organoaluminum compounds demonstrate a high catalytic activity. For example, the system Co(acac) 2 -AlEt 3 yields the following mix of dimers: 1 -62%, 3 -29%, hexacyclic dimers 9%. The direction of this reaction can be changed through adding phosphines. When PPh 3 is used, Binor-S forms rather selectively. If diphos is substituted for PPh 3 , hexacyclic dimer 8 is generated almost quantitatively. This system has a unique feature: it catalyzes not only NBD dimerization, but also its co-dimerization with NBN. The yield of the co-dimer is 42%.
Co(acac) 3 -AlEt 2 Cl-Diphos + Rhodium complexes probably manifest the most versatile synthetic possibilities as compared with other catalysts of NBD cyclodimerization. They can catalyze all the above-described types of cycloaddition and also give new dimers 15-17.
Nickel catalysts are the best for [2π+2π]-cycloaddition. The ratio of isomers depends on the type of ligands in nickel complex, temperature, solvent and other conditions [18].
NBD cyclodimerization was found to begin with an induction period, the length whereof depends on the complex type and temperature. The reason for this phenomenon is that the NBD-nickel catalyst forms through substituting initial ligands, and this process is different for each system (3)(4)(5) [19].  Direction (3) is characteristic of complexes with ligands stable in the free state: Ni(CO) 4 , Ni(1,5-Ñ 8 Í 12 ) 2 and Ni(CH 2 =CHCN) 2 . The second direction takes place for compounds containing unstable moieties. They can decompose through ligand coupling (4) or intramolecular disproportionation (5).
Intermediates Ni(NBD) n were separated at different stages of the catalytic process with various initial nickel compounds. The molar ratio of NBD to Ni proved to be 2.07 (±0.03) in all these cases, which testifies that one the same catalyst forms in situ in all the systems: it is an equilibrium mix of complexes Ni(NBD) 2 and Ni(NBD) 3 at a ratio of 12÷14 to 1.
Kinetic studies showed that the formation rate of dimers 1, 3 and 8 (with taking into consideration that Ni complexes with different number of NBD ligands exist simultaneously in the reaction mass) are described by the following kinetic equations: where k 1 , k 3 , è k 8 are observed cyclodimerization rate constants.
Even though these equations retain the same form for different solvents, the type of medium influences the ratio of products and the general rate of the process. It should be emphasized, that dimers 1, 3 and 8 have different kinetic orders with respect to NBD, which is quite unusual for concurrent formation of isomers with a similar structure.
Observed rate constants k 1 , k 3 , and k 8 are in a satisfactory agreement with the Arrhenius equation. Activation parameters are presented in Table 1.
Phosphine nickel complexes where at least two places are blocked with ligands are not active in NBD cyclodimerization. When the ratio of PR 3 to Ni is equimolar, dimers 1, 3 and 8 are formed in accordance with a common kinetic equation, which is right for various phosphines: The mutual ratio of resulting cyclodimers of NBD depends on the steric characteristics of the phosphine, which probably influence the mutual orientation of NBD molecules coordinated in the complex. The proportion of dimer 1 is growing with an increase in the conical angle of phosphine.
The catalytic action mechanism of nickel complexes was scrutinized.
An NBD molecule can be either a η 2 -or a η 4coordinated ligand. A bidentate coordination of NBD is thermodynamically more favorable than a monodentate one. The difference between these two states equals to the homoconjugation energy of double bonds (about 8.0 kJ/mol), which can contribute to the stability of the resultant complex. Kinetic data testify that three different π-complexes of Ni and NBD are formed (19)(20)(21); there is an equilibrium between them (Fig. 2) [20].
When considering the stereochemistry of cycload- dition, we must take into account not only the coordination type of an NBD molecule on a nickel atom (monodentate or chelate), but also its exo/endo-orientation. When an NBD is coordinated monodentately, its exo-coordination is preferable from both sterical and thermodynamic points of view. A stereochemical analysis of various possible orientations of NBDmoieties in complex Ni(NBD) 4 testifies that four NBD ligands can only be exo-coordinated there. A dimer molecule is formed already in the intermediate π-complex, and the NBD-nickel system is a matrix predetermining the structure of the product. Oxidative addition of coordinated NBD molecules to the Ni center results in the creation of metallocyclic intermediates. Decomposition of the latter is the limiting stage. Monodentate NBD-molecules remaining in the coordination sphere after the complex has decomposed transform into chelate ligands and stabilize Ni(0).
The results of our studies of the kinetics and the mechanism of the process allow us to forecast the behavior of the system in various conditions. For examples, if one coordination vacancy is blocked on the Ni atom (the ratio of PR 3 to Ni = 1:1), kinetic regularities of the process change. Use of phosphines with large conical angles is favorable for the more compact exo-coordination of NBD molecules and, consequently, results in an increased content of dimer 1.
The difference in kinetic orders with respect to NBD in the above equations testifies that the formation rate of isomer 1 depends on NBD concentration more strongly than that of isomer 3. Analysis of the activation parameters of the process indicates that the yield of exo-trans-endo isomer 3 is growing dramatically as temperature is increasing (Fig.3).
Isomer 1 is formed quantitatively in the temperature range from 10 to 30°C in non-polar media at NBD concentrations above 3 mol/l. As reaction temperature is growing and NBD concentration is de-creasing, the relative amount of exo-trans-endo isomer is increasing; its yield reaches 95% at 100°C and an NBD concentration of 0.5 mol/l. Virtually any required ratio of 1 and 3 can be achieved through varying these parameters.
The process rate decreases in the presence of phosphines. Besides, the selectivity cannot any longer be controlled through NBD concentration in phosphinecontaining systems [21].

Co-dimerization of NBD with unsaturated compounds.
Co-dimerization with NBD is a very versatile and prospective way of synthesizing polycyclic compounds. It often is an alternative to diene syntheses based on cyclopentadiene. Sometimes co-dimerization is preferable.
The range of substrates suitable for catalytic codimerization with NBD is rather narrow. It is obvi-  To form a mixed hetero-substrate complex, the second co-monomer and NBD must have comparable abilities to coordinate on the metal atom. Ankynes, a range of alkenes with highly-reactive double bonds, and some oxygen-and nitrogen-containing compounds are suitable for this process. Co-dimerization of NBD with unsaturated compounds results in products of [2π+2π] and [2π+4π]-cycloaddition; each direction of such a reaction can give a range of spatial and optical isomers.
[2π+2π]-cycloadducts retain an intra-cyclic norbornene double bound. They can be used for further transformations, for example in methathesis oligomerizaitoin or polymerization. When NBD codimerize with some dienes or alkynes, the resulting product has several double bonds with different reactivity, which is rather attractive for manufacturing of special-purpose rubbers.
All the above-mentioned confirms that such compounds can be used as semiproducts in medicine, microelectronics, perfumery etc. The range of application of NBD co-dimerization-based processes is limited for some reasons. They are possible homodimerization of NBD or the second substrate; formation of side products through consecutive cycloaddition to both the double bonds of NBD; formation of isomers of all kinds: skeletal, spatial (exo/endo, cis/trans, enantio) ones. Hence it is extremely important to develop highly active and selective catalytic systems for the aforementioned processes.
Even though unsaturated NBD-derivatives usually react with alkenes with acceptor substituents to give [2π+4π]-cycloadducts, sometimes we can observe [2π+2π]-cycloaddition with a high regio-and even stereo-selectivity. These reactions usually proceed through the unsubstituted double bound of NBD, and substituents in locations 2 and 3 can favor one direction of the reaction or the other, exo-or endo-.
Unsubstituted NBD usually reacts to give [2π+4π]adducts. These reactions proceed far easier in the pres-ence of transition metals, and their stereoselectivity changes, too. For example, if the thermal process mostly gives endo-adducts, a catalyst promotes formation of exo-cyclic compounds [23][24] (reaction 10 and Table 2   Para-cycloadduct 26 is the major product for electron-acceptor substituents, ortho-isomer 27 is for electron-donor ones, while a CN-moiety promotes formation of metha-isomer 28. So, the type of the substituent in diene and dienophile can affect the selectivity of [2π+4π]-cycloaddition. A substituent in position 7 in the NBD-ring also influences the regio-selectivity of the process. An extremely high exo-selectivity of cycloaddition is characteristic of such reactions, independent of the type of the substituent. However, syn/anti-isomerism becomes possible in this case. The yield of anti-isomer is growing with an increase in the electronegativity of substituent 7 [25] (reaction 16 and  Table 4 Selectivity of reaction (16) vs. type of substituent Y in position 7 of the NBD ring. Changes in reaction temperature or substitution of phosphite for phosphine influence selectivity of such cycloadditions weakly.
These data agree completely with the results of quantum-chemical ab initio calculations of 7 substi-tuted NBDs carried out in the basis STO-3G , [26][27] A bimetallic system capable of catalyzing reaction between CO 2 and NBD under mild conditions was synthesized from bis-π-allyl nickel and cationic π-allyl palladium complexes [31]. cat.
As a result of the reaction, isomeric pentacyclic lactones form.
In such systems even non-activated alkynes react Table 5 Effect of the substituent type in alkyne on regularities of reaction (19)  First publications have appeared recently where enantio-selective cyclocodimerization of NBD and alkynes is described. The system Co(acac) 2 -Et 2 AlCl with chiral diphosphines was used as a catalyst. The regioselectivity of these processes with the participation of substi-tuted NBDs is low (50÷70%), the enantioselectivity reaches 85÷90%.
Very interesting examples of intramolecular cycloaddition with the participation of 2-alkenyl-substituted NBD with various lengths of the hydrocarbon bridge were described [28] (reaction 21): The aklyne moiety has two possible ways of attachig to NBD. However, only direction 30 takes 33 increases up to 68 %, and product 34 virtually ceases forming [33].
When system Co(acac) 3 -dppe-Et 2 AllCl in ben-zene is used, NBD can dimerize not only with 1,3butadiene, but with its derivatives as well [34] (reaction 24): Lautens and his colleagues performed this reaction enantioselectively recently [35]. Co(acac) 2 or Co(acac) 3 can be used for the initial cobalt compound, and a chiral diphosphine can be applied as the ligand. The best results were obtained for Rdiphenylphosphinepropane (R-Propos), where the selectivity to ee-isomer exceeded 70% (reaction 25 and Table 6):  Table 6 Effect of the phosphine type on the reaction yield and stereoselectivity in reaction 25.  (26) rather low yield of [4π+2π+2π]-adducts. In addition to them, NBD dimers and products of polymerization form.
The codimerization processes involving NBD have a range of qualitative analogies. Intramolecular type of co-dimer formation implies formation of heteroligand π-complexes of nickel; the ability of the substrate to co-coordinate has to be between that of the mono-and bidentated NBD molecule. One must know these regularities to obtain individual products with a high yield and selectivity.

Catalytic allylation of NBD.
Catalytic allylation of the strained double bond in NBD-derivatives is a promising way to obtain compounds with unique structures. It was first described by M. Catellani and G. Chiusoli in 1979 [36]. This method was substantially developed in studies by [37][38][39]. A series of NBD-derivatives incorporating two or more double bonds with different reactivity can be obtained in one technological stage through this reaction. The various reactivity of the double bonds is rather valuable when these compound are used as comonomers for special-purpose synthetic rubbers.
Allylic esters of organic acids are the source of allyl moieties. The most unusual feature of this reaction is the character of addition, which can be not only linear (the conventional type), but cyclic as well, or even proceed with breaking a C-C bond.
Catalytic system -Ni(acac) 2 -AlEt 3 -P(OR) 3 is very active and sufficiently selective when used to obtain exo-methylenecyclobutane derivatives. Sterical features of substrates incorporating exo-substituents in positions 5 and 6 with respect to the intracyclic double bond do not exert a substantial effect on the rate of cycloaddition. Quite the opposite, the reactions slows down considerably in the case of endo-substituents. The structure of allylating agent (allylic ester) is not decisive. For example, you can use not only allylacetate (AA), but allylpropoinate, allylbenzoate and allylformate as well. However, you will not be able to attain a high conversion of reagents due to decomposition of the catalytic system in the last case.
The shares of individual isomers among the oligomeric products depend of temperature, the type and structure of the organophosphous ligand. At 25°C, major products are methylenecyclobutane derivatives.
The structure of substituents in location 7 of the NBD molecule exerts little influence on co-oligomerization catalysts. However, substituents in positions 2 and 3 of the NBD-moiety can block this reaction route (reaction 28) Catalytic allylation of NBD and NBD derivatives is an extremely complicated multiparameter process, which proceeds unconventionally. Different factors affect its rate, yield, turnover, and selectivity.
The kinetics and mechanism of the process were scrutinized through a complex of physical and chemical methods for catalytic system Ni(C 3 H 5 ) 2 (P(O-iC 3 H 7 ) 3 ) n . Its results indicate that there are equilib-rium stages like: NiL n NiL n-1 + L Complexes with different number of phosphite ligands were detected through 31 P NMR in model systems and in conditions of a catalytic process. Phosphite gets re-esterified by acetic acid, one of the reaction products. This leads to the formation of a mix of phosphonates and deactivates the catalytic system. That NBD molecules can coordinate in the complex in different ways makes the process more complicated, and so does the successive allylation of two double bonds (Fig. 4).
Analysis of the kinetic data let us assert that individual products 45-47 form due to the presence of nickel complexes with different number of phosphite ligands in the catalytic system (Fig. 5). The experimental data allow us to propose a mechanism of catalytic allylation of NBD (Fig. 6). In accordance with it, complexes all NiL n OAc dominate in the reaction media. An NBD molecule co-ordinates on a nickel atom, which induces η 3 -η 1 -isomerization of the allyl ligand, and then inserts into the η 1 -allylmetal bond. Then, depending on «n», cyclization proceeds in one route or another. It finishes with a βhydride transfer, formation of products and regeneration of NiL n (a quick oxidative addition of an allylacetate molecule to NiL n from solution). When n = 1, NBD coordinates chelately in the complex, which gives rise to product 47 that has a nortricyclene structure.  Hydrid transfer is probably the limiting step of the process. This step proceeds with the participation of a β-carbon, which was confirmed through using model system C 3 D 5 OCOCD 3 -NBD (Fig. 7).
... Acetic acid destabilizes the catalytic system. This is why the number of catalyst turnovers does not exceed 150-200 in a static reactor.
The lifetime of the catalyst and the number of its turnovers can be increased up to 2000 through use of zeolites adsorbing acetic acid selectively.
Applying the regularities found, we can obtain product 45 with a yield up to 85% and an output of 350 g/(l·hour) at 80°C and a L/Ni ratio of 3. The product 46 forms with a selectivity of 95% and an output of 220 g/(l·hour) at 25°C and a P/Ni ratio of 2.
The product 47 can be obtained with a selectivity up to 70% at a L/Ni ratio equal to and a temperature of 50 to 70°C.
A very interesting situation exists when a double excess of allylacetate over NBD is used. Secondary allylation of norbornene-type compounds 45 and 46 yields a wide range of isomeric products (reaction 30). Structures and proportions of double allylation products depends on conditions (the temperature and P/Ni ratio). Allylation of a double bound in NBD is several times faster than the second stage, so the kinetic curves depicting accumulation of products 45 and 46 have pronounced maximums (Fig. 8).
Hence NBD allylation has a general character. It can be used as a promising method for obtaining rare polycyclic hydrocarbons.

Conclusion
Synthetic possibilities of NBD and its derivatives are extremely versatile. However, this positive factor gives rise to problems caused by simultaneous realization of several reactions in a single system. Largescale use of NBD is substantially limited through difficulties connected with separation and analysis of isomeric products, problems of sensible utilization of reagents and insufficient catalyst activity. Metal-complex catalysts provide unique opportunities for obtaining various polycyclic hydrocarbons. Use of these complexes is the most promising way of developing this synthetic direction. Transition metals possess the basic ability to influence selectivity to any degree; however, a thorough information about the mechanism of their action is necessary to implement these possibilities. Unfortunately, relevant data are badly scarce in literature.
To find a decent place for NBD among principal substrates in organic and petrochemical synthesis we need deeply understand the basis of NBD-involving processes, which in turn requires us to use a complex of various physical and chemical methods and synthetic techniques, as well as scrutinize the kinetic regularities in the relevant systems.