Three Models of the Metallic Bond in Poly(sulfur nitride)


Poly(sulfur nitride) – also known as polythiazyl and (SN)x ; abbreviated as PSN in this web page – was discovered serendipitiously by the English chemist F.P. Burt in 1910 . He was purifying a sample of tetrasulfur tetranitride by vacuum sublimation , passing its vapors through a filter of silver wool to remove the sulfur impurities . The silver (actually silver sulfide , Ag2S) catalyzed the decomposition of the S4N4 vapor to thiazyl free radicals , which dimerized to S2N2 . This unstable intermediate polymerized to PSN in Burt's vacuum sublimation apparatus . The PSN formed a thin blue film and a thicker layer with a metallic , bronze luster . Burt also prepared bronze PSN crystals by using a quartz wool filter instead of the silver . He apparently suspected these products were metallic , but never confirmed his idea at the time .

PSN was examined in the 1950s by the German chemist M. Becke-Goehring , who investigated it together with numerous other sulfur-nitrogen compounds . She made the first electrical conductivity measurements on PSN , describing it as a semiconductor with an ambient conductivity of the pressed powder of only 100 ohm-1-cm-1 . Two decades later electrical measurements on very pure single crystals of PSN , which had a much higher conductivity , reaffirmed the metallic nature of the material . A definitive review of PSN with many references , including details of its historical development , is provided by Labes , Love , and Nichols . The references are listed below , at the end of the web page .

This essay is not intended as a review of PSN research ; rather , the results of my own study of (SN)x are presented here . They have been distilled down into three models of understanding the metallic bond in PSN from a chemistry perspective . The Valence Bond Theory is used to describe the S–N covalent bonds – the stronger , lower energy “spine” of the molecules – in a simple , pictorial manner . Then the higher energy metallic bond is created above , and surrounding , the covalent skeleton by the continuous overlapping of unhybridized , native atomic orbitals , to form the “polymerized molecular orbital” containing the unused , “leftover” valence electrons . This polymerized molecular orbital is a crystal orbital (abbreviated as XO) and is the metallic bond , or conduction band , in the material .

The Valence Bond description of a compound is closely correlated with its known physical and chemical properties , so the crystal structure and other physical and chemical features of PSN will be reviewed before the first model of its metallic bond is presented .


Some Physical and Chemical Properties of Poly(sulfur nitride)


I've written about PSN elsewhere in Chemexplore , so it's convenient for me to “copy and paste” the relevant sections describing the material :

PSN consists of acicular crystals with a golden color and metallic luster when viewed from above , or dark blue when the ends of the crystals are observed . They are somewhat soft and malleable , and can be rolled into thin sheets . When left exposed to the air and humidity , PSN crystals tarnish and become dull like conventional base metals . PSN reacts with one-electron oxidizers such as bromine to form black “salts” that are also highly conductive . PSN is a rather unstable material . It sublimes in vacuo at about 135 C , and explodes in air at around 240 C (a thermal instability typical of most , if not all sulfur-nitrogen compounds) , and will also explode if it's strongly compressed .

X-ray diffraction shows that the crystal structure of PSN consists of long chains of alternating sulfur and nitrogen atoms in a crenellated , or “all-Z” pattern :

The sketch immediately above was copied from the Wikipedia web page , “Polythiazyl”. I thank the author of this sketch , and Wikipedia , for implied permission to reproduce it on this web page . Underlined blue hyperlinks can be clicked when online to download the PDF , HTML , or GIF file , which will open in a new window) .

The PSN bond lengths and angles in the sketch above were determined by X-ray diffraction . Somewhat different results were obtained by a French researcher , M. Boudeulle , using an electron diffraction technique . Presumably X-rays are less interactive with the sulfur and nitrogen atoms in PSN than are electrons , so the assumption is that the X-ray derived data are more accurate and representative than are those from electron diffraction . A convenient tabulation of both data sets is provided by Mikulski and co-workers . As we'll see later in this web page , the bond angles will be of critical importance in the Valence Bond analysis of PSN's electronic structure . The bond lengths also provide valuable insight into the type of bonding in the material .

The long S–N chains are bonded sideways in a parallel manner to form fibers (which have been observed by scanning electron microscopy) , which in turn are packed into fiber bundles . The acicular PSN crystals are formed from these fiber bundles .

The ambient electrical conductivity of very pure , well-formed PSN crystals is about 4000 ohm-1-cm-1, and the compound becomes superconducting at Tc = 0.26 K . It has a small ambient magnetic susceptibility (~ 9 x 10-6 emu/mol) ; this paramagnetism was found to be temperature independent . Such a “t.i.p.” or Pauli paramagnetism is typical of the conventional , eg. elementary , metals and their alloys , and of synthetic metals . It originates from the random spins of the free electrons above the Fermi level , EF , in the material's metallic bond .

PSN has an inverse electrical conductivity–temperature relationship as measured along (or parallel to) the crystal lengths , but has a direct relationship when measured across (or perpendicular to) them :

This sketch was adapted from the graph of the electrical resistivity of poly(sulfur nitride) by Grant and co-workers . My thanks to the copyright holder .

The ambient (273-295 K) resistivity of PSN in the sketch is ~ 5 x 10-4 ohm-cm , which translates into an electrical conductivity of around 2000 ohm-1-cm-1 [the resistivity multiplier , 104 (shown) , was inverted to the conductivity multiplier , 10-4 (not shown) , when the graph was flipped upside down] . The conductivity of PSN is highly dependent on the chemical purity and degree of perfection of the crystal examined .

The nearly linear inverse conductivity–temperature relationship displayed by PSN indicates that it is a True Metal , i.e. has a nodeless XO as its metallic bond . Of lesser significance is the secondary trace , that of the direct conductivity–temperature relationship across (or perpendicular to) the component fibers . Such a direct relationship is typical , even diagnostic of the nodal XO metallic bond found in Pseudometals , such as the semiconductors . The “nodes” in this case are undoubtedly the physical gaps between the fibers , which constitute the energy gaps that the conduction electrons must “jump” or “tunnel through” in order to travel through the PSN crystals in a sideways manner . Such jumping or tunneling through the gaps , or nodes , requires energy , which must be supplied by the environment (or the human researcher making the measurement) . The more energy supplied , the more electrons can cross the fibres , and the greater its conductivity ; hence , a direct conductivity–temperature relationship is observed .

In terms of the electronic structure of PSN , the inverse parallel relationship is the important one , because it applies to the movement of the conduction electrons along and over the chains of sulfur and nitrogen atoms via the principal metallic bond . The inverse relationship tells us that the XO forming the metallic bond will be nodeless , a very important orbital property that must be accounted for in the models we'll be looking at . The native , unhybridized atomic orbitals comprising the XO must all overlap in such a way that the XO is nodeless along at least one crystal axis . Then , PSN can behave like a True Metal , as we know it does from the electrical conductivity measurements .

Three models of the metallic bond in PSN can be proposed . The component orbitals and electrons can be derived from either of the sulfur or nitrogen atoms alone , or from both of them together . Based on the known physical and chemical properties of PSN the Valence Bond theory is used to establish the electronic configuration of the sigma covalent bonds in the PSN chains ; then , any leftover valence electrons will be located in energetically accessible frontier (LUMO) orbitals . These latter orbitals are then polymerized over the S–N spines to provide a nodeless XO , which should be suitable as the metallic bond in the polymer .

We'll see that each of the three models has both positive and negative attributes ; selecting the “best” model will then be a matter of further experimental verification or refutation .


The First Model : Sulfur Atoms Only Sigma XO


Chemical intuition suggests that sulfur should be more metallic than nitrogen , since as a rule of thumb as the elements in a family on the Periodic Table become heavier they tend to become more metallic in nature . So intuitively sulfur should be our first choice as the source of the metallic bond in PSN .

In the “sulfur only” sigma XO model the sulfur and nitrogen atoms both have a trigonal planar hybridization , sp2 + pz . The nitrogen atom (2s2 2p3) would resemble that one in pyridine , with two sigma bonds , a lone pair of nonbonding electrons , and a single electron in the 2pz native , unhybridized orbital . The sulfur atoms (3s2 3p4) would be similar , but in this case the sixth valence electron is “leftover” and is promoted to a frontier orbital . This would occur because of resonance between the nominally sulfur(II) and sulfur(IV) atoms alternating in the (SN)x chains :

In this picture there simply isn't any room for the sulfur atoms' sixth valence electrons in the resonating pi electron system . They must be promoted into higher energy frontier (LUMO) orbitals . I suggested the sulfur 4s orbitals as a possible location for them . The continuous overlapping of these 4s native orbitals along the PSN chains would certainly produce an excellent nodeless sigma XO metallic bond for the material . The s orbitals are quite voluminous , and should provide a strong overlap throughout the sigma XO .

The sulfur–sulfur distances in PSN are a rather long 2.789 (sketch above) . Compare this to the S–S bond length , 2.06 , in elementary S8 sulfur (2.04 in S2Cl2 ; 2.05 in organic disulfides) . Assuming that the sulfurs’ 4s AOs have a metallic radius of 2.35 – similar to that of the potassium atom , whose valence electron is also 4s1 – the length of the overlap of the 4s AOs along the sulfur-sulfur separations would be (2 x 2.35 ) – 2.789 = 1.911 , which is 68 % of the S–S bond length . There could thus be a substantial overlap of the 4s AOs over the sulfurs to produce a strong sigma XO along the (SN)x polymer chains . This will be nodeless , and would be satisfactory as the metallic bond in PSN . The physical and chemical properties of PSN (softness , malleability , rapid tarnishing , one-electron oxidation) mentioned above are also consistent with a 4s1 electronic configuration for the sulfur atoms in poly(sulfur nitride) . PSN would thus be isoelectronic and chemically (and even somewhat physically !) similar to potassium metal , but is much less reactive than it , of course .

The Valence Bond sketch below summarizes the electronic structure of PSN according to the “sulfur atoms only” sigma XO model :

The best feature of this model is that the sulfur 4s orbitals would be excellent for the metallic bond , being very voluminous and more than able to span the long S–S distances and efficiently overlap along and over the polymer spines . However , two considerations give us pause for concern . First , the N–S–N bond angle is much too small (106) for trigonal planar hybridization ; the ideal trigonal angle is 120 . There could be a small amount of steric compression of the N–S–N bond angle by the voluminous sulfur lone pair (if trigonal) , but probably not enough to account for the 14 reduction .

Second , there might not be enough system energy available to promote the sulfur sixth valence electrons into the 4s orbitals . MacDiarmid and co-workers have suggested that these electrons might be located in the S–N antibonding molecular orbitals (ABMOs) . However , such ABMOs are nodal in nature , and would therefore be unsuitable for use in PSN's metallic bond , although it might be in the S–N pi XO at a lower energy level in their model (this will be examined in the next section) .

Sulfur might use its normally vacant 4 s-p orbitals in two bonding scenarios . The first is in the molecular metal TTF–TCNQ , where aromatization of the TTF1+ component forces its seventh sulfur electron out of the ring and promotes it up into a frontier (LUMO) orbital , most likely the sulfur 4s AO . TTF–TCNQ is a True Metal above 58 K ; below that temperature its electrical conductivity “crashes” and it rapidly becomes virtually an insulator . Possibly the 4s electrons move into ABMOs below 58 K , making it a Pseudometal (semiconductor) . In any case , there is no aromatization in PSN , so that type of system energy is unavailable for promotion of a sulfur 3pz valence electron into the 4s orbital .

The second scenario where sulfur might combine its vacant 4 s-p orbitals in hybridization with the 3 s-p orbitals and electrons is in highly exoergic reactions , particularly in fluorinations . Sulfur burns quite exothermically (288.9 Kcal/mol) in fluorine to form sulfur hexafluoride , SF6 , and some of that reaction energy can be used in hybridization of the sulfur 3,4 s-p orbitals , with the six valence electrons . I've proposed in another web page that sulfur could form a sp5 octahedral hybrid orbital for SF6 from the 3s , 3px,y,z , and 4px,y orbitals . The sp5 hybrid orbital reflects the fact that sulfur is a p-block element , and would therefore be expected to hybridize mostly its native p orbitals for use in sigma covalent bonds . However , once again the sulfur 4s orbital might be inaccessible for PSN , which is formed at room temperature – not in a high temperature exothermic reaction – by the polymerization of S2N2 .

Let's move on to the next model of PSN's metallic bond , in which both the sulfur and nitrogen atoms are participating electronically .


The Second Model : Sulfur and Nitrogen Atoms' Bilayer Pi XO


The Valence Bond description of the electronic conditions of the sulfur and nitrogen atoms is similar in this model to that described for the “sulfur atoms only” model above . In this case , though , the sulfur sixth valence electrons remain in the 3pz native orbitals instead of being promoted up into a frontier orbital . The sulfur atoms will then have an electronic configuration similar to that one in the thiophene molecule (trigonal planar sp2 + pz = sa + sb + lpxy + 3pz2) . The nitrogen 2pz1 and the sulfur 3pz2 native orbitals then overlap continuously along and over the S–N spines to form a nodeless bilayer pi XO , which should be suitable as PSN's metallic bond :

In this model the sulfur 3pz2 pairs of electrons are in the lower energy part of the XO , below the Fermi level (EF) , while the nitrogen 2pz1 singlet electrons are all above EF . This sort of a bilayer XO is common in metallic Transition metal oxides , most notably in rhenium trioxide , ReO3 , which I consider to have the best example (GIF , 34 KB) of such a metallic bond .

This is an attractive model for PSN's metallic bond ; the pi bonds should readily form between the sulfurs and nitrogens (the S–N bond length is ~ 1.61 , which isn't excessive) ; there should be a large population of free electrons above EF resulting in a good electrical conductivity in the material ; and the pi XO is nodeless , thereby resulting in a True Metal behaviour in PSN , which is actually observed . However , as with the previous model , the sulfur atoms in this second model also require a trigonal planar hybridization with a N–S–N bond angle of about 120 . Unfortunately this much smaller bond angle , 106 , is almost certainly that of the sp3 tetrahedral hybridization (which has the ideal – uncompressed – angle of 109 28') .

Confronting this N–S–N bond angle reality head-on , we arrive at the third model of the metallic bond in PSN , that of the “nitrogen atoms only” scenario . It's the most unorthodox of the three models , with startling implications , so I deliberately left it as the last model to be examined .


The Third Model : Nitrogen Atoms Only Sigma XO


In the third model of PSN's metallic bond the sulfur atoms have a tetrahedral configuration , similar to that of sulfur in the familiar sulfur(II) molecules hydrogen sulfide , the mercaptans (thiols) , sulfides (thioethers) , and disulfides . The S–S–S bond angles are 106 in the S8 rings of elemental sulfur ; those sulfur atoms certainly have a tetrahedral sp3 configuration . Four of the six valence electrons in tetrahedral sulfur are in two lone pairs which are non-bonding and don't participate in the metallic bond in PSN . The other two are in the S–N covalent bonds .

The nitrogen atoms have a trigonal configuration , with S–N–S bond angles at 119.9 , almost the ideal trigonal angle of 120 . Let's suppose nitrogen's configuration is trigonal planar , as in the previous two models . Since the sulfurs are electronically inert linking atoms , we must have a direct –N–N– metallic bond , a nodeless pi XO composed of continuously overlapping 2pz1 native orbitals along the polymer spines . In theory , that would be a satisfactory metallic bond for PSN ; but in practice , it could probably never form .

The N–N distance in PSN is 2.576 (sketch above) , as measured by X-ray diffraction . The N–N bond length in hydrazine , H2N–NH2 , is 1.45 ; in the nitrogen molecule , it's 1.0976 . The N=N double bond length in the hypothetical PSN metallic bond would be approximately the average of the hydrazine single bond and nitrogen triple bond lengths , i.e. 1.27 , which is half of the actual N–N distance in PSN . A conventional N=N pi double bond couldn't span such a long distance . How about an unconventional N=N sigma bond ?

In this picture the nitrogens have a trigonal pyramid configuration , which are s [native , unhybridized] + p3 , the latter being the actual hybrid orbital per Valence Bond Theory :

Two examples of compounds in which the s + p3 electronic configuration might be found are cadmium iodide and titanium disulfide . Both have a layered crystal structure in which the metal atoms are octahedrally coordinated by the nonmetal atoms , which in turn are pyramidally coordinated to them . CdI2 and TiS2 have the cadmium iodide crystal structure , which consists of alternating planar sheets of anion–cation–anion layers :

Blue spheres : Cd2+ cations (octahedral) ; green spheres : iodide anions (pyramidal) . A section of one sheet is illustrated .

Chemical bonding in cadmium iodide is both ionic and coordinate covalent . In the latter case trigonal pyramid hybridization of the 5p native orbitals of the iodide anions (5s2 5p6) provides three sigma bonding lobes , each with a lone pair of electrons . These pairs are donated to the electrophilic cadmium cations , which receive them in an empty octahedral hybrid orbital , possibly sp3d2 .

Titanium disulfide also has the cadmium iodide crystal structure , with octahedral titanium and trigonal pyramid sulfur . Bonding in TiS2 is metallic (Ti–Ti) and coordinate covalent (S-->Ti) . The titanium atoms are formally zerovalent – not ionic or covalent Ti(IV) – in this extraordinary compound . Its suggested electronic structure , per Valence Bond Theory , reflects these atomic coordinations :

Titanium disulfide is a metallic solid with a golden-yellow color and a bright metallic luster . It's a fair electrical conductor , with an ambient conductivity of about 1400 ohm-1-cm-1, and it has an inverse electrical conductivity–temperature relationship from room temperature to near Absolute Zero . TiS2 has a flaky morphology like graphite , its platelets consisting of three-layer “sandwiches” of sulfur–titanium–sulfur . It forms many molecular complexes with intercalated reagents . Only electron donors such as lithium metal , organic amines , and amides are intercalated in TiS2 ; electrophiles aren't known to intercalate between its layers . This is consistent with the picture of the empty sulfur 3s native , unhybridized orbitals , which can accept donated electrons ; obviously , they have none to transfer to electrophiles .

Titanium disulfide is discussed at greater length in another web page . The s + p3 electronic configuration for the iodide anions and sulfur atoms in CdI2 and TiS2 , respectively , seems to be a satisfactory rationalization for their distinctive layered crystal structures .

Returning to the “nitrogen atoms only” model for PSN's metallic bond , the nitrogen 2s1 orbitals would be isoelectronic with the 2s1 orbitals in lithium metal , which has a metallic radius of 1.57 . Assuming this value for the hypothetical nitrogen 2s1 , there would be an overlap of (2 x 1.57 ) – 2.576 = 0.564 , which is 22% of the N–N distance in the PSN chains . While not as great an overlap as with the “sulfur atoms only” 4s sigma XO model (68%) , it should nevertheless be sufficient for the formation of a N–N sigma XO metallic bond along the PSN chains .

The S–N bond length of 1.61 in PSN supports the “nitrogen atoms only” model . Mikulski and co-workers , in discussing the S–N bond length in S2N2 , stated ,

“........ suggesting a bond order of approximately 1.5 for all of the equivalent S–N bonds . The S–N single bond length is expected to be approximately 1.74 and that for the S–N double bond is approximately 1.54 . Their average , 1.64 , expected for a bond order of ca. 1.5 , is remarkably close to that found experimentally (1.654 )” [Mikulski et al. , p. 6361 ; Gritsan and co-workers] .

The bond order in PSN is thus 1.5 , as in S2N2 (1.654 ) and S4N4 (1.616 ) . In the “sulfur atoms only” model , the system has a bond order of 2.5 (sigma S–N + pi S–N + metallic bond S–S) . Note that the metallic bond is in effect a crystal-wide (XO) collectivity of resonating one-electron bonds , which adds a 0.5 bond order to the system .

In the sulfur–nitrogen bilayer metallic bond model , the bond order would again be 2.5 by the same reasoning . By reverse reasoning , if the S–N bond length in PSN , 1.61 , indicates a bond order of 1.5 , then there are only the sigma S–N covalent bonds and the N–N metallic bond . There is no S–N resonating pi electron system . The sulfur atoms aren't involved in the electronic structure ; they are only inert linking atoms .

NB : I think that's generally the electronic configuration in S2N2 , S4N4 , and PSN . Their primary strong bonds are the S–N sigma covalent bonds , which form the skeletal frames of the molecules . Then there is some other secondary weak resonating electron system at a higher energy level , and physically covering the covalent bonds , providing the extra 0.5 bond order . In PSN , it's the N–N metallic bond XO ; in S4N4 it could be resonating N–N one-electron bonds . In S2N2 it might be the two nitrogen 2pz1 singlet electrons , resonating in a nitrogen 2pz–sulfur 4pz pi MO over the four-atom ring .

The N–N sigma XO metallic bond would be entirely consistent with the known physical and chemical properties of poly(sulfur nitride) . And finally , this last nitrogen model also respects the known S–N–S (trigonal) and N–S–N (tetrahedral) bond angles , so it seems to be the most preferable of the three models of PSN's metallic bond reviewed in this web page . But what a peculiar , counterintuitive model nonetheless ! How can we test its validity by physical and chemical experiments ?

Physically , the electronic nature of the metallic bond orbitals could be analysed spectroscopically . ARPES (angle resolved photoelectron spectroscopy , also known as ARUPS , angle resolved ultraviolet photoelectron spectroscopy) measures the energy distribution of the valence shell electrons in atoms . It should be able to differentiate between the sigma XO and pi XO models , and possibly also between the sulfur and nitrogen types of sigma XO as well . An ARPES analysis of poly(sulfur nitride) should be most revealing .

Chemically , it might be possible to design and synthesize analogous polymers in which alternate electronically inert linking atoms are substituted for the sulfurs . The “nitrogen atoms only” model for PSN's metallic bond will be supported if the resulting polymer is metallic . In the following final section of this web page a picture of such a PSN analogue polymer is presented .


Poly(dimethylcarbonitride) , [(CH3)2CN]x : a New Metallic Polymer ?


In the “nitrogen atoms only” model the sulfurs have a tetrahedral sp3 configuration and are electronically inert linking atoms . It might be possible to replace them in the –S–N– chain with other isoelectronic atoms or radicals . Tetrahedral carbon , –(R1R2)C– , comes immediately to mind . The repeating unit would then be –(R1R2)C–N– , derived from the imine monomer (R1R2)C=N . Its dimer , so to speak , would be an azine , which is well-known type of organic compound . Azines are readily synthesized from the condensation of two equivalents of an aldehyde or ketone with hydrazine , to produce an aldazine or ketazine , respectively :

A wide variety of aldehydes and ketones have been condensed with hydrazine in the preparation of azines . Four hydrazine reagents have been used in the syntheses : anhydrous hydrazine (actually 95–98% pure) ; hydrazine hydrate (the azeotrope with water , containing 64% NH2–NH2) ; the crystalline salt , hydrazine hydrochloride ; and the most recent version , hydrazinium carboxylate . The carbonyl and hydrazine components have been combined neat or in a solvent ; in the former case , with anhydrous hydrazine , the reaction can be quite exothermic . Note : hydrazine is a very toxic , potentially dangerous (explosive) chemical ; be aware of its hazards and handle it carefully , observing the necessary safety procedures .

The laboratory preparation of azines is usually quite straightforward . Azines are produced on a commercial (i.e. multi-ton) scale as a captive intermediate – used internally , and not marketed externally – in the manufacture of hydrazine , an important industrial chemical . The European conglomerate Pechiney-Ugine-Kuhlmann operates such an azine/hydrazine production unit in southern France ; it's referred to as the “PUK process” :

This image was copied from the Wikipedia web page , “Pechiney-Ugine-Kuhlmann Process”. I thank the author of this sketch , and Wikipedia , for implied permission to reproduce it on this web page . The sketch shows the use of MEK (methyl ethyl ketone) in a later alternate process, but in earlier practice acetone was used to produce acetone azine . The central intermediate was 3,3-dimethyloxaziridine , which was ammoniated to acetone hydrazone . The acetone azine was hydrolyzed (aqueous acid , then neutralized with alkali) to hydrazine hydrate , with the recovery of it and the acetone in a fractional distillation column .

The simplest azine , formaldazine , CH2=N–N=CH2 , is derived from formaldehyde ; however , it's a gas that readily polymerizes at room temperature . Neureiter prepared it in 50% yield as a polymer , which was pyrolyzed to the gas ; this product was condensed in a Dry Ice trap , where it formed colorless or white crystals , m.p. –48 C , soluble in polar solvents . When allowed to warm to room temperature , the formaldazine spontaneously polymerized , first to a viscous liquid , then to a white solid .

Around 1970 Horne and Norrish prepared formaldazine gas and studied its decomposition under ultraviolet irradiation :

“In the case of formaldazine the condensation reaction resulted in the formation of a white polymer which was dried in a desiccator and pumped at 50 C on a vacuum line to remove traces of water and formaldehyde . The solid was depolymerized at approximately 220 C and formaldazine was purified by trap to trap distillation and stored at –196 C . Operations with the gas were performed at less than 30 Torr pressure to suppress wall polymerization . Acetaldazine and dimethylketazine were purified by trap to trap distillation to give colourless liquids and stored over calcium oxide at room temperature” (Horne and Norrish , from their p. 303) .

Monomeric phenylformaldazine (1-phenyl-2,3-diaza-1,3-butadiene) , Ph–CH=N–N=CH2 , has been prepared by Japanese chemists . It was initially a pale yellow oil which quickly changed into a pale yellow powder , m.p. 130-135 C (dec) , that was used in polymerization studies .

The best known and most practical aliphatic azine for investigation as a monomer in the hypothetical polymerization reaction is undoubtedly acetone azine (dimethylketazine , b.p. 131C) . It would provide the electronically inert –(CH3)2C– spacers for the polymer . As its name suggests , it's readily prepared from acetone (2 eq.) and hydrazine (1 eq.) . Horne and Norrish studied the flash photolysis (UV light) of formaldazine , acetaldazine , and dimethylketazine , concluding that “N–N fission is the major process under all experimental conditions” ; the resulting free radicals then disproportionate and decompose in various complex secondary reactions . Forty years later another study of the UV decomposition of acetone azine by Brinton and Chang again found a complex mix of reaction products indicating an initial homolysis of the molecule's N–N bond (the products all had only one nitrogen atom each) followed by various rearrangements of the imine free radicals .

Azines are reasonably stable compounds with no indication of any explosive nature . They decompose to various products when pyrolysed ; in fact , this is a lesser-known olefin synthesis :

“The azines on heating alone are decomposed with loss of nitrogen to give an unsaturated compound , R–CH=N–N=CH–R -----> N2 + R–CH=CH–R , and this reaction is sometimes used for preparing substituted ethylenes” (Sidgwick , p. 522) .

More recent studies of the pyrolysis of azines have revealed a somewhat more complex situation ; in some cases phenyl–CH=N– fragments (from benzalazine) can expel hydrogen atoms to provide benzonitrile as the main product .

The challenge in this proposed synthesis of poly(dimethylcarbonitride) is finding a suitable catalyst that first homolytically cleaves the N–N bond of the acetone azine monomer , then catalyzes the polymerization of the resulting imine free radicals into [(CH3)2CN]x , all the while avoiding the disproportionation and decomposition of these highly reactive and somewhat unstable molecules .

Several complex Transition metal-based (ruthenium , tungsten) catalysts have been synthesized and studied by Grubbs and co-workers and Schrock and co-workers ; they have the ability , relevant in this context , of cleaving olefin bonds and then catalyzing the polymerization of the free radical intermediates . For example , cyclooctatetriene was converted into the “pre-metallic” polymer polyacetylene , (CH)x , using Grubbs's catalysts :

While Grubbs' catalysts are of interest in this proposed cleavage/polymerization of azines , competing reactions might prevent the [(CH3)2CN]x formation . The catalyst might merely promote the polymerization of the uncleaved azine into its corresponding nonmetallic polymer , as monomeric formaldazine does so readily even uncatalyzed . Another strong possibility might be the ROMP (ring opening metathesis polymerization) reaction rearranging the azine molecule into nitrogen gas and an olefin . The imine bonds would be cleaved by the catalyst , not the N–N bond , leaving the =N–N= section behind as the very thermodynamically stable nitrogen molecule . The (CH3)2C= fragments would then join together as (CH3)2C=C(CH3)2 , 2,3-dimethyl-2-butene (b.p. 73 C) , in the case of acetone azine . Generalized , this would make a neat olefin synthesis but it wouldn't be very helpful in the preparation of [(CH3)2CN]x .

A variety of Transition metal coordination compounds have been developed during the past couple of decades that are capable of cleaving the nitrogen molecule ; they accomplish this remarkable feat by forming very strong metal–nitrogen triple bonds . The bound nitrogen atom is then referred to as a nitrido ligand . Protonolysis of the nitrido atom releases ammonia . In effect , molecular nitrogen from the atmosphere has been hydrogenated (often at room temperature with these new catalysts) to ammonia . Interestingly , an azine-like intermediate has been proposed for the formation of the metal–nitrido complex : 2 L3M + N2 ------> L3M=N–N=ML3 ------> 2 L3M~N , where ~ (tilda symbol) represents a triple bond . The ligand , L , is always a very bulky alkyl/aryl group , which prevents the L3M complex from dimerizing by forming a L3M~ML3 triple bond . Various Transition metals have been used in these complexes ; one that is particularly reactive toward nitrogen is trivalent molybdenum , Mo(III) .

Thus , one of Professor Schrock's “nitrogen fixation catalysts” could conceivably cleave the N–N bond of acetone azine : 2 L3Mo + (CH3)2C=N–N=C(CH3)2 ------> 2 L3Mo=N.=C(CH3)2 [note the singlet electron on the nitrogen atom] . Would the N.=C(CH3)2 part of the complex subsequently polymerize into the metallic polymer , or would it remain firmly bonded to the molybdenum ?

Ziegler-Natta catalysts could also be tried with acetone azine . They have been successfully used in nitrogen fixation experiments to cleave the N–N bond in nitrogen molecules , resulting in the production of hydrazine and ammonia . Low valent Transition metal cations , notably Ti(III) and V(II) , are typically the active species in these Ziegler-Natta type of catalysts . Are they also forming transient L3M=N–N=ML3 types of intermediates with the nitrogen ? Since Ziegler-Natta catalysts are used in large volumes in the polymerization of olefins , and in particular isoprene (the latter to form all-cis-polyisoprene , synthetic rubber) , and since acetone azine somewhat resembles isoprene , the catalyst might merely promote the rapid polymerization of intact , i.e. uncleaved azine molecules to an uninteresting waxy – and completely nonmetallic – resin .

Finding the magic “Catalyst X” to accomplish the N–N bond cleavage of the azine and subsequent polymerization of the (CH3)2C=N imine free radicals would be a challenging project , but with the tantalizing goal of synthesizing a novel metallic polymer it could be a very rewarding one . The new polymer would be a fascinating material in its own right ; and its successful creation would provide considerable credibility to the “nitrogen atoms only” model for the metallic bond in poly(sulfur nitride) .


References , Notes , and Comments


Labes , Love , and Nichols : M.M. Labes , P. Love , and L.F. Nichols , “Polysulfur Nitride – A Metallic , Superconducting Polymer”, Chem. Rev. 79 (1) , pp. 1-15 (1979) ; see also : R.T. Oakley , “Cyclic and Heterocyclic Thiazenes”, Prog. Inorg.Chem. 36 , pp. 299-391 , S.J. Lippard (ed.) , John Wiley , New York , 1988 ; G.B. Street and W.D. Gill , “The Chemistry and Physics of Polythiazyl , (SN)x , and Polythiazyl Halides”, pp. 301-326 in Molecular Metals , W.E. Hatfield (ed.) , Plenum Press , New York (1979) .

electrical measurements : V.V. Walatka Jr. , M.M. Labes , and J.H. Perlstein , “Polysufur Nitride – A One-Dimensional Chain with a Metallic Ground State”, Phys. Rev.Lett. 31 (18) , pp. 1139-1142 (1973) ; C-H. Hsu and M.M. Labes , “Electrical Conductivity of Polysulfur Nitride”, J. Chem. Phys. 61 (11) , pp. 4640-4645 (1974) .

more accurate and representative : Professor MacDiarmid (MacDiarmid and co-workers , below) commented ,

“By using more reliable single-crystal X-ray diffraction methods , we recently (Mikulski and co-workers) demonstrated that these data [electron diffraction data from M. Boudeulle et al.] are not correct” (p. 64) .

Mikulski and co-workers : C.M. Mikulski et al. , “Synthesis and Structure of Metallic Polymeric Sulfur Nitride, (SN)x , and Its Precursor , Disulfur Dinitride , S2N2”, J. Amer. Chem. Soc. 97 (22) , pp. 6358-6363 (1975) ; see their Table III , p. 6362 , for PSN X-ray diffraction data .

becomes superconducting : R.L. Greene , G.B. Street and L.J. Suter , “Superconductivity in Polysulfur Nitride (SN)x”, Phys. Rev. Lett. 34 (10) , pp. 577-579 (1975) . The Tc of PSN might be significantly increased if its metallic bond was coupled to the spin rgime of a strongly antiferromagnetic (AFM) compound . Then , AFM induction would suppress the Pauli paramgnetism in the metallic bond free electrons above EF , the Fermi level , in effect creating a “pre-Cooper pair” state in them . Less cooling would be required to cause these “pre-Cooper pairs” to fully condense into strongly bound “real” Cooper pairs ; that is , Tc will have been substantially increased from its normal value of 0.26 K in virgin PSN . The condensation and polymerization of S2N2 vapor onto the surface of a variety of AFM compounds should produce an epitaxial thin film of PSN on them . These composite heterostructures could then be tested for their electrical conductivity and superconductivity properties .

Grant and co-workers : P.M. Grant et al. , Comparison of the Physical Properties of Polysulfur Nitride , (SN)x , to Related Organic Polymer Systems and (TTF) (TCNQ)”, Mol. Cryst. Liq. Cryst. 32 (1) , pp. 171-176 (1976) [PDF , 387 KB] ; see esp. Figure 1 , p. 172 (graph of the resistivity of PSN) . This research paper was downloaded from Paul Grant's website , which offers free PDF copies of many of his scientific publications .

metallic radius : A.F. Holleman , E. Wiberg , and N. Wiberg , Inorganic Chemistry , 1st ed. Engl. , Academic Press , San Diego (CA) , 2001 ; Table 21 ,Metal Atom Radii, p. 136 .

MacDiarmid and co-workers : A.G. MacDiarmid et al. , “Synthesis and Selected Properties of  Polymeric Sulfur Nitride (Polythiazyl) , (SN)x”, Ch. 6 , pp. 63-72 in Inorganic Compounds with Unusual Properties , R.B. King (ed.) , Adv. Chem. Series 150 , American Chemical Society , Washington , D.C. (1976) .

Gritsan and co-workers : N.P. Gritsan et al. , “Matrix Isolation and Computational Study of the Photochemistry of 1,3,2,4-Benzodithiadiazine”, J. Phys. Chem. A 111 (5) , pp. 817-824 (2007) [PDF , 300 KB] . These researchers calculated S–N and S=N bond lengths of 1.71 and 1.56 , respectively , in the title compound (Figure 8 , p. 823) :

ARPES : A.F. Santander-Syro , “Introductory Lectures on Angle-Resolved Photoemission Spectroscopy (ARPES) and Its Application to the Experimental Study of the Electronic Structure of Solids” [PDF , 1742 KB] ; ARPES Research Group , University of British Columbia [web page] ; A. Damascelli , “Probing the Electronic Structure of Complex Systems by ARPES”, Physica Scripta T109 , pp. 61-74 (2004) [PDF , 1859 KB] ; Shen Laboratory, Stanford University , CA [web page] ; EE290f - Lecture 19 : “Angle Resolved Photoemission and Non-ARPES”, given by Dr. Eli Rotenberg , ALS/LBNL [YouTube video , FLV , run time 1:07:35 ; a high speed Internet connection is recommended , if not mandatory , for viewing this video !] .

preparation of azines : A.C. Day and M.C. Whiting , “Acetone Hydrazone”, Org. Synth. Coll. Vol. 6 , p. 10-12 (1988) [PDF , 134 KB] ; H.H. Hatt , “Methylhydrazine Sulfate”, Org. Synth. Coll. Vol. 2 , pp. 395-397 (1943) [PDF , 119 KB] ; E.C. Horning et al. , “Preparation of 3-Methyl-5-Aryl-2-Cyclohexen-1-ones”, J. Amer. Chem. Soc. 68 (3), pp. 384-387 (1946) ; idem. , “Aromatization Studies. V. Synthesis of Alkylanilines from Alkylcyclohexenones”, 69 (8) , pp. 1907-1908 (1947) ; idem. , “Aromatization Studies. VI. Aminobiphenyls and Naphthylamines”, 70 (1) , pp. 288-289 (1948) ; L.D. Frederickson Jr. , “An Infrared Study of the C=N Stretching Vibration in Azine Derivatives of Aldehydes and Ketones”, Analyt. Chem. 36 (7) , pp. 1349-1355 (1964) ; H.M. Nanjundaswamy and M.A. Pasha , “Rapid , Chemoselective and Facile Synthesis of Azines by Hydrazine / I2”, Synth. Commun. 37 (19) , pp. 3417-3420 (2007) ; J. Safari and S. Gandomi-Ravandi , “Highly Efficient Practical Procedure for the Synthesis of Azine Derivatives Under Solvent-Free Conditions”, Synth. Commun. 41 (5) , pp. 645-651 (2011) .

The preparation of imines (Schiff bases) is quite similar to that of azines . See for example , R.B. Moffett , “N-Methyl-1,2-Diphenylethylamine and Hydrochloride”, Org. Synth. Coll. Vol. 4 , p. 605-608 (1963) [PDF , 129 KB] ; R.B. Moffett and W.M. Hoehn , “Analgesics . II. The Grignard Reaction with Schiff Bases”, J. Amer. Chem. Soc. 69 (7), pp. 1792-1794 (1947) ; K.N. Campbell , A.H. Sommers , and B.K. Campbell , “The Preparation of Unsymmetrical Secondary Aliphatic Amines”, J. Amer. Chem. Soc. 66 (1) , pp. 82-84 (1944) ; K.N. Campbell et al. , “The Reaction of Grignard Reagents with Schiff Bases”, J. Amer. Chem. Soc. 70 (11), pp. 3868-3870 (1948) .

As with unsubstituted or less-substituted azines , the imines are very reactive and rapidly polymerize at room temperature . Campbell and co-workers (JACS 66) noted ,

“The aldimines obtained in this work were water-white when freshly distilled ; they are unstable , and polymerize on standing , and should be used a few hours after distillation” (p. 82) .

Anderson similarly found that N-alkylmethylenimines , eg. N-methylmethylenimine , CH2=NCH3 (b.p. ~ –35 C) , very rapidly polymerized to a “resinous or wax-like polymer” when warmed to room temperature : J.L. Anderson , “Preparation of Methylenimines”, U.S. Patent 2,729,679 (January 3rd , 1956) [PDF , 168 KB . Note : this file can be opened only with Adobe Acrobat Reader v. 6 or later . If desired , this application can be downloaded for free from] .

The general conclusion from all of this research with azines and imines is that the less-substituted azines are strongly susceptible to spontaneous polymerization at room temperature and should probably be avoided in any investigation into the formation of the hypothetical metallic polymer . The tetramethyl-substituted acetone azine is less liable to spontaneous polymerization at room temperature than those azines with one or more CH functionalties .

hydrazinium carboxylate : B. Lee et al. , “Synthesis of Azines in Solid State : Reactivity of Solid Hydrazine with Aldehydes and Ketones”, Org. Lett. 13 (24) , pp. 6386-6389 (2011) . The reagent , H3N+–NH–COO-, was prepared by the reaction of hydrazine with CO2 in supercritical carbon dioxide . Perhaps it could be more conveniently prepared by thoroughly carbonating a solution of 95% hydrazine in isopropanol (b.p. 82 C) , then removing the solvent by rotary evaporation in vacuo . Isopropyl alcohol forms an azeotrope (b.p. 80 C , with 88% alcohol , 12% water) ; evaporation of this azeotrope would ensure the resulting solid , crystalline product is dry .

Neureiter : N.P. Neureiter , “Monomeric Formaldazine – Synthesis of 1,3,4-Thiadiazolidine – A New Heterocycle”, J. Amer. Chem. Soc. 81 (11), p. 2910 (1959) . DOI . I don't usually provide DOI (digital object identifier , i.e. online abstract) links in my web pages , since their articles are pay-per-view , and I don't link to pay-per-view anything in Chemexplore ; however , ACS often provides a scan of the first page of the article in their DOIs , and in this rare case all of Neureiter's brief paper is in the scan , so you can read his entire report about formaldazine in the DOI .

Horne and Norrish : D. G. Horne and R. G. W. Norrish , “The Photolysis of Acyclic Azines and the Electronic Spectra of R1R2CN Radicals”, Proc. Roy. Soc. Lond. Ser. A 315 (1522) , pp. 301-322 (1970) . A study of the UV photolysis of benzalazine in solution : R.W. Binkley , “The Photochemistry of Unsaturated Nitrogen-Containing Compounds . I . The Irradiation of Benzalazine”, J. Org. Chem. 33 (6) , pp. 2311-2315(1968) . When the photolysis was carried out in cyclohexane and with benzophenone as a sensitizer , benzalazine provided exclusively benzonitrile in an 85% yield (Table I , p. 2312 ; and see the sketch below) .

phenylformaldazine : A. Hashidzume et al. , “Preparation and Polymerization of Benzaldehyde Formaldehyde Azine (1-Phenyl-2,3-diaza-1,3-butadiene)”, Macromolecules 33 (7) , pp. 2397-2402 (2000) . As with the less-substituted azines and imines mentioned above , this compound also seems to be quite susceptible to a certain degree of spontaneous polymerization at room temperature .

Brinton and Chang : R.K. Brinton and S. Chang , “The Photolysis of Dimethylketazine Vapor”, Ber. Bunsen. Physik. Chem. 72 (2) , pp. 217-221 (2010) .

Sidgwick : N.V. Sidgwick , The Organic Chemistry of Nitrogen , 3rd ed. , Clarendon Press , Oxford (UK) , 1966 . Azines are briefly discussed on pp. 521-522 .

pyrolysis of azines : H.E. Zimmerman and S. Somasekhara , The Mechanism of the Thermal Decomposition Reaction of Azines”, J. Amer. Chem. Soc. 82 (22), pp. 5865-5873 (1960) ; C.G. Overberger and P-K. Chien , “The Thermal Decomposition of Benzalazines in Solution”, J. Amer. Chem. Soc. 82 (22), pp. 5874-5876 (1960) . The pyrolysis (and photolysis ; see Binkley's paper in Horne and Norrish , above) might be represented as follows :

Binkley's research suggests another approach to [(CH3)2CN]x . Acetone azine could be photolysed with UV light in solvents of varying polarity , from relatively nonpolar (cyclohexane , benzene , ethyl ether) to very polar (propylene carbonate) , in a quartz vessel . A Pyrex™ vessel usually filters out the energetic part of the UV light required to break the N–N bonds ; it can still be used , but benzophenone must be included in the irradiated solution as a sensitizer . Under these conditions benzalazine will homolyse and expel its methylene protons , resulting in the formation of benzonitrile in high yields . However , as there are no methylene protons in acetone azine , perhaps the imine free radicals , (CH3)2C=N , will be reasonably stable at STP and will polymerize on the vessel walls to [(CH3)2CN]x . The radicals might also be stabilized by a judicious choice of a particular solvent . Binkley found that the less polar solvents , cyclohexane and benzene , gave much higher yields of benzonitrile than the fairly polar methanol .


The N–N bond in acetone azine might be cleaved catalytically by the reagent tris(4-bromophenyl)aminium hexachloroantimonate , which is a radical cation salt . [I've written about “tris” in another web page] :

Tris is a mild oxidizer and can extract a p electron from an olefinic bond . It might be able to effect a one-electron oxidation of one or other of the nitrogen atoms in acetone azine , thereby leading to its homolysis to the imine free radicals , which would then hopefully polymerize to [(CH3)2CN]x . Tris can be prepared from the corresponding amine ; it's also commercially available .

Grubbs and co-workers : L.K. Johnson , S.C. Virgil , R.H. Grubbs , and J.W. Ziller , “Facile Tungsten Alkylidene Synthesis : Alkylidene Transfer from a Phosphorane to a Tungsten Imido Complex”, J. Amer. Chem. Soc. 112 (13), pp. 5384-5385 (1990) ; C.B. Gorman , E.J. Ginsburg , and R.H. Grubbs , “Soluble , Highly Conjugated Derivatives of Polyacetylene from the Ring-Opening Metathesis Polymerization of Monosubstituted Cyclooctatetraenes : Synthesis and the Relationship between Polymer Structure and Physical Properties”, J. Amer. Chem. Soc. 115 (4), pp. 1397-1409 (1993) ; L.K. Johnson et al. , “Alkylidene Transfer from Phosphoranes to Tungsten( IV) Imido Complexes”, J. Amer. Chem. Soc. 115 (18), pp. 8167-8177 (1993) .

Schrock and co-workers : R.R. Schrock et al. , “Preparation and Reactivity of Several Alkylidene Complexes of the Type W(CHR')( N-2,6-C6H3-i-Pr2)(OR)2 and Related Tungstacyclobutane Complexes . Controlling Metathesis Activity through the Choice of Alkoxide Ligand”, J. Amer. Chem. Soc. 110 (5) , pp. 1423-1435 (1988) . See also : B.L. Langsdorf , X. Zhou , and M.C. Lonergan , “Kinetic Study of the Ring-Opening Metathesis Polymerization of Ionically Functionalized Cyclooctatetraenes”, Macromolecules 34 (8) , pp. 2450-2458 (2001) .

Protonolysis of the nitrido : D.G.H. Hetterscheid , B.S. Hanna , and R.R. Schrock ,“Molybdenum Triamidoamine Systems . Reactions Involving Dihydrogen Relevant to Catalytic Reduction of Dinitrogen”, Inorg. Chem. 48 (17) , pp. 8569-8577 (2009) ; A.E. Shilov , “Intermediate Complexes in Chemical and Biological Nitrogen Fixation”, Pure & Appl. Chem. 64 (10) , pp. 1409-1420 (1992) [PDF , 899 KB] ; see esp. Table 2 , “Protonation of Bridging Dinitrogen Complexes”, p. 1417 .

azine-like intermediate : C.E. LaPlaza et al. , “Dinitrogen Cleavage by Three-Coordinate Molybdenum(III) Complexes: Mechanistic and Structural Data ”, J. Amer. Chem. Soc. 118 (36) , pp. 8623-8638 (1996) ; Scheme 1 , p. 8625 ; J.J. Curley et al. , “Shining Light on Dinitrogen Cleavage : Structural Features , Redox Chemistry , and Photochemistry of the Key Intermediate Bridging Dinitrogen Complex”, J. Amer. Chem. Soc. 130 (29) , pp. 9394-9405 (2008) ; Scheme 1 , p. 9395 ; M.B. O'Donoghue et al. , “ “Fixation” of Dinitrogen by Molybdenum and the Formation of a Trigonal Planar Iron-Tris[molybdenum(dinitrogen)] Complex”, J. Amer. Chem. Soc. 119 (11) , pp. 2753-2754 (1997) .

trivalent molybdenum : C.E. LaPlaza and C.C. Cummins , “Dinitrogen Cleavage by a Three-Coordinate Molybdenum(III) Complex”, Science 268 (5212) , pp. 861-863 (1995) ; D.V. Yandulov and R.R. Schrock , “Reduction of Dinitrogen to Ammonia at a Well-Protected Reaction Site in a Molybdenum Triamidoamine Complex”, J. Amer. Chem. Soc. 124 (22) , pp. 6252-6253 (2002) . The Mo in trigonal planar Mo(III) complexes , when sterically protected from dimerization , bonds very strongly and selectively to pnictide atoms . L3Mo will cleave the N–N (not N–O) bond in nitrous oxide : L3Mo + N=N=O ------> L3M~N + L3M–N=O : C.E. LaPlaza et al. , “Cleavage of the Nitrous Oxide NN Bond by a Three-Coordinate Molybdenum(III) Complex”, J. Amer. Chem. Soc. 117 (17) , pp. 4999-5000 (1995) . L3Mo will also strongly bond to the phosphorus atoms from white phosphorus : C.C. Cummins , “Reductive Cleavage and Related Reactions Leading to Molybdenum–Element Multiple Bonds : New Pathways Offered by Three-Coordinate Molybdenum(III)”, Chem. Commun. 1998 (17) , pp. 1777-1786 ; L3Mo~P complexes are discussed on pp. 1781-1782 . This chemistry suggests that the N–N bond in acetone azine would indeed be cleaved by the L3Mo Mo(III) reagents developed by Professors Schrock and Cummins and their co-workers : 2 L3Mo + (CH3)2C=N–N=C(CH3)2 ----------> 2 L3Mo=N.=C(CH3)2 , as mentioned above . However , the N.=C(CH3)2 residue may be too strongly bound to the Mo to further polymerize into a metallic polymer .

nitrogen fixation experiments : M.E. Vol'pin with V.B. Shur , “Nitrogen Fixation by Transition Metal Compounds : Results and Prospects”, J. Organometallic Chem. 200 (1) , pp. 319-334 (1980) . The systems listed in Table 1 , “Nitrogen-fixing Systems”, p. 322 , are strongly reducing in nature . Many of the reagents listed in Table 2 , “Catalytic Reduction of N2 by Aluminum and Lithium Hydrides”, p. 326 , resemble Ziegler-Natta catalysts for olefin polymerization .

in particular isoprene : H. Tucker and S.E. Horne Jr. , Elastomers , Synthetic (Polyisoprene)”, pp. 582-592 in the Kirk-Othmer Encyclopedia of Chemical Technology , 3rd edition , Vol. 8 , M. Grayson and D. Eckroth (eds.) , John Wiley , New York (1979) . A formula for the polymerization of isoprene to poly(cis-1,4-isoprene) using a typical Ziegler-Natta catalyst : hexane , 900 parts (w/w) + isoprene , 100 parts + tri-isobutyl-aluminum , 2.11 parts + titanium tetrachloride , varied from 0.8 to 1.1 mole ratio , Ti/Al ; temperature rising from 0–50 C , 2–4 hrs. reaction time for a 90% conversion .

Professor H. Shirakawa's group in Japan polymerized acetylene to the “pre-metallic” polyacetylene back in the early 1970s using a Ziegler-Natta catalyst . Virgin polyacetylene has a lustrous , silvery , metallic appearance , but is essentially an insulator . It must be doped , preferably with a one-electron oxidizer such as iodine , in order to exhibit any appreciable electrical conductivity . The catalyst system they used was a mixture of tetrabutoxytitanium and triethylaluminum [Chem. Abs. 80 , 108933x (1974)] , with an Al/Ti mole ratio of 34 . The critical catalyst concentration was 3 mmoles or higher . In Japanese Patent no. 73 32,581 , Oct. 6th , 1973 [Chem. Abs. 81 , 26223x (1974)] , a similar catalyst formula was revealed : toluene , 30 ml + tetrabutoxytitanium , 1.7 ml + triethylaluminum , 2.7 ml ; this was stated to be a 4 : 1 molar ratio of Al/Ti . This catalyst system provided mostly cis-polyacetylene at 78 C ; roughly 50 : 50 cis : trans at room temperature ; and almost pure trans-polyacetylene at 150 C .



A New Look at the Chemical Bonding in S2N2 and S4N4

[ Index Page ] [ Contact ]