How I Discovered A Platinum Mine One Day

(And Lost It The Next)


Life is full of adventures , some good , some bad . Whether good or bad , though , I always try to learn something from them . It's especially nice if I can learn some chemistry from my experiences . As one of my chemistry professors once advised me , “Live and learn ; it's much better than just living”. Of course , he's an academic , so I suppose he may be somewhat biased in his opinion . But at least it's a positive philosophy .

I've had many educational chemistry adventures , one of the more memorable ones occurring when I was employed in R&D in the asbestos industry of the Eastern Townships of Quebec , Canada , back in the early 1980s . The following story is a sort of cautionary tale , nicely illustrating the old English adage , “The best laid schemes of mice and men oft go astray” (paraphrasing the great Scottish poet Robbie Burns) . I relate it here because it warns us against overconfidence and haste , and teaches us to be careful and diligent in our research . And there's some interesting chemistry in the story , too .

Our Research Center was divided into several laboratories with their associated research groups . Mine was the asbestos fiber section and laboratory . A wide variety of scientific and technical investigations were carried out in the Research Center at that time , all concerned with one aspect or other of the Quebec asbestos industry . A major emphasis was placed by our management on finding new uses for the immense quantities of asbestos mill tailings produced by mines in the Eastern Townships .


Asbestos Tailings


Tailings from the tail-end of the ore milling process consist of a fine to coarse granular mineral material , gray in color , gritty and dusty . These tailings are the inevitable waste by-product of asbestos fiber production , accumulating in immense volumes which are piled up in miniature mountains near the mills . Here are several photos of asbestos tailings in the Eastern Townships :


These pictures are from the Internet ; I don't recall the sources of the top two . The last one is from a web page of the Centre de Technologie Minérale et Plasturgie Inc. (CTMP) . My thanks to all the copyright holders of these images . [The references are listed at the end of this web page . Underlined blue hyperlinks can be clicked when online to download the PDF or HTML file , which will open in a new window] .

And for American readers , here's a photo of the tailings dump at Lowell , Vermont (Vermont Asbestos Group mine) , which I found on an EPA web page :


Lowell is located near Newport , in the “Northeast Kingdom” of Vermont ; its underlying serpentine geology is a southern continuation of that in the Eastern Townships of Quebec , along the northeastern spine of the Appalachian Mountains .

The minerology of the tailings is quite complex . They consist primarily of serpentinite minerals (the interested reader is referred to the Wikipedia web page Serpentinite for more information on its geochemistry) . The host rock of chrysotile asbestos fiber is serpentine , which has an opaque , milky green color . In fact , substantial quantities of the chemically related semi-precious gemstone jade were discovered and mined in , and exported from the Cassiar asbestos mine (now closed) in northern British Columbia . Both serpentine and its fibrous counterpart chrysotile asbestos have the same ideal empirical formula , Mg3Si2O5(OH)4 . The hydroxide anions are ionically bonded to the magnesium cations , which cover the SiO skeleton . These hydroxide anions are very slightly soluble in water , making serpentine and chrysotile faintly alkaline in a water suspension . The fibrous mineral form of magnesium hydroxide , Mg(OH)2 [familiar as the laxative “Milk of Magnesia”] , is called brucite , and is similarly faintly alkaline . Because of their mildly alkaline nature the serpentinite minerals are geologically classified as ultrabasic or ultramafic .

Transition metals mainly iron are a significant impurity in serpentine and chrysotile from the Eastern Townships . Interestingly , chrysotile asbestos mined in Arizona is embedded in a limestone (calcium carbonate) matrix , and has virtually no iron impurity ; it's a sparkling snow-white fiber , quite unlike the drab gray Quebec asbestos . Iron occurs both in the fiber itself , where ferrous cations , Fe2+, can substitute to a minor extent for the magnesium cations , and as a magnetite impurity , Fe3O4 , adhering tightly to the fiber bundles (spelks and spicules) . These ferrous cation and magnetite impurities impart the typical bluish-gray color to both the Townships chrysotile and the mill tailings shown in the pictures above . When asbestos and tailings samples are calcined in a muffle furnace at , say , ~ 1000 ºC , atmospheric oxygen oxidizes the Fe2+ to Fe3+, and the fiber and tailings samples become a light tan and dark brown color , respectively (magnetite is concentrated in the tailings in the milling process , so they have a much higher iron content than the fiber) .

Nickel and cobalt are also found in serpentine , although at much lower concentrations than iron . The appreciable concentration of magnesium in serpentine (26.31% per the stoichiometry) has made asbestos tailings of considerable interest for a long time as a possible source of magnesium metal and magnesium chemicals . Many elaborate schemes have been devised to extract magnesium from tailings . In the course of these studies detailed chemical analyses have been carried out on them . Transition metal impurities in serpentine minerals have also been examined from an environmental point of view , as they are regarded as a possible toxic component of the otherwise innocuous magnesium minerals . Chemical analyses of Albanian and Spanish serpentine soils have revealed surprising levels of Transition metals in them . For example , in the former serpentine 3865 mg/Kg of chromium , 476 mg/Kg of cobalt , 3579 mg/Kg of nickel , and 1107 mg/Kg of copper were found . In the Spanish serpentine samples the concentrations were 2236 , 2163 , 169 , and 57 mg/Kg , respectively .

In the attempted extraction of the magnesium from tailings by an aqueous method (for example , with a hot mineral acid solution) most of these Transition metal components of the serpentine are co-extracted with the magnesium salt and are very difficult to separate from it . In a second approach the tailings are finely ground and are calcined in a furnace at ~ 1000 ºC . In this heat treatment the hydroxide anions are converted into water molecules , and the calcined tailings then consist of an anhydrous magnesium silicate , plus the Transition metal impurities , of course .

The calcined tailings , which are a dark brown powder resembling rust , can be fused at a very high temperature (usually in an arc furnace at ~ 2000 ºC) with an alkaline reagent under reducing conditions . The alkali dissolves the silicate part of the mineral , forming a water-soluble glass from which the magnesium hydroxide component can be extracted . An added reducing agent (usually carbon , in the form of graphite or coke) chemically reduces the Transition metal components to their elementary metal form . These metal globules are denser than the molten silicate phase , and settle to the bottom of the refractory crucible which is usually made of graphite where they agglomerate into larger nodules , or oval-shaped “buttons” . These metal buttons can be recovered after the alkali silicate glass has been removed from the crucible .

I know of two such alkali fusion processes , both of which produced these Transition metal buttons as a recoverable by-product . In the first process the tailings were fused with sand (silica) and I suspect , an alkali of some sort or other to form a molten mass that was spun into a mineral wool material . Ferronickel buttons containing 56% nickel were co-produced in this process . I had the opportunity to inspect a sample of this serpentine-based mineral wool at the CANMET research laboratory in Ottawa , courtesy of its inventor , Albert Winer . It was an attractive , glassy wool with a pretty , pale blue color . Unfortunately , it proved to be susceptible to degradation by moisture . With the benefit of over three decades of hindsight , I would suggest that this hydrolysis shortcoming might be corrected by the addition of ~ 5% or so of boron oxide , B2O3 , to the melt ; this by analogy to the borosilicate glasses Pyrex™ and Vycor™ , which are much superior in their thermal and chemical resistance properties to ordinary soda-lime [sodium calcium silicate] glass .

The second alkali fusion process was carried out in the High Temperature laboratory at our Research Center . A mixture of the calcined tailings , carbon , and anhydrous sodium carbonate (“soda ash”) was melted by an electric arc struck between its graphite crucible container , as one electrode , and a large carbon rod in the center of the crucible , immersed in the reaction mixture . The huge electric current required for the fusion was produced by an arc welding apparatus . I observed the fusion process on one occasion , and it was very impressive , almost scary in fact , with the thunderous roar , clouds of smoke , and fiery sparks : modern-day alchemy ! Townships asbestos tailings consist almost exactly of 90% of a silaceous phase (the magnesium silicate) and 10% of a feruginous phase (the Transition metal components , mostly iron) . The main product of this fusion process was therefore a sodium magnesium silicate , whose degree of water solubility could be controlled by adjusting the sodium : magnesium ratio . The minor product , considered at the time to be little more than a laboratory curiosity by the High Temperature researchers , was the feruginous phase : the button-like iron nodules recovered from the bottom of the graphite crucible . Our story really begins with those unremarkable iron buttons from the High Temperature lab .


Platinum Group Metals in Ultrabasic Rock such as Serpentine


I don't recall three decades after the event when I first examined the iron buttons from the alkali fusion , nor under what circumstances . More than a decade before that I had carried out several thermit reactions with Transition metal oxides . The manganese dioxide reduction with aluminum powder (a violent reaction !) produced a dark brownish-black oval button of impure manganese metal , and the iron buttons from the calcined tailings reminded me of it . The High Temperature group knew that there was nickel and maybe cobalt in the buttons , but apart from that there was little interest in them , and they were set aside as oddities .

I recalled that the two main producers of platinum group metals (mainly platinum and palladium) were South Africa and Russia . Canada is another source of platinum and palladium ; they occur to a minor extent with nickel in northern Ontario (the Sudbury region) . Interestingly , South Africa and Russia are also producers of chrysotile asbestos , and the South African platinum occurs in ultrabasic host rock , as does chrysotile . Similarly , asbestos has been found (and for a while , mined) in the Sudbury region , although to a much lesser extent than in the Townships of Quebec . There seems to be a geological connection between the platinum group metals , ultrabasic minerals , serpentine , and chrysotile asbestos , reminiscent of the geological connection between diamonds and kimberlite . However , the connection is a very general one , since South Africa , Russia , and Canada are blessed with a great wealth of many industrial minerals from which are derived metals and nonmetal commodities such as asbestos .

The chemical connection of the platinum group metals with serpentine is more apparent . Iron , cobalt , and nickel are significant Transition metal components of serpentine , originating from the olivine , fayalite , and magnetite found in serpentinite minerals (Serpentinite) . The platinum group elements are merely heavier members of the Group VIII / 8910 Transition metal family :

This annotated Periodic Table was scanned from the Alfa-Aesar catalog , “Research Chemicals , Metals and Materials” [a very useful reference] . My thanks to the copyright holder . For a much larger , more readable Periodic Table you can download this GIF image (279 KB) . It's a scan of a Periodic Table handed out at my High School chemistry final examination way back in June , 1965 ! While lacking the more recently-discovered Transuranium elements , it nevertheless provides all the electron configurations of the 92 “standard” elements , so it may be a helpful reference for modern-day chemistry students .

So it's not unreasonable to think that there might actually be measurable amounts of platinum group metals in serpentine , and therefore in the calcined tailings . If so , they would have been chemically reduced to their elementary form in the fusion process and concentrated in the iron buttons .


The Platinum Mine is Discovered


The above reasoning prompted me to discuss the possibility of detecting platinum group metals (commonly called “PGMs”) in the iron buttons with our Research Center management . Fortunately they were very open-minded and receptive to novel ideas , so they readily agreed with my suggestion that one of the buttons be sent to a geology firm I knew of in Ottawa , for a chemical analysis of platinum and maybe other PGMs [almost a decade earlier I had worked as a geochemical and geophysical field assistant for one of the partner-owners of this firm when I was a college student] . This was done , and we awaited the results with considerable anticipation .

The analytical technique used on the iron button was that of cupellation , also commonly referred to as a fire assay . The button was first weighed very accurately , then cooled in liquid nitrogen ; in its superchilled condition it was very brittle , and was pulverized to a fine powder in a mortar . A weighed quantity of the powder was mixed with reagent grade lead metal , and the two were fused together . The lead alloy button was then placed on a bone ash (or MgO) cupel and heated in a furnace . The purpose of the lead was to act as a metallic solvent , to extract any noble metals from the base metals (mainly iron) . Cupellation is a common analytical procedure for gold and silver in ores , and presumably for PGMs as well . In the Amazon River basin in Brazil the native miners mix dirt and sand dug from the river banks with mercury , which acts in a similar manner as the lead in cupellation to dissolve the very fine particles of gold that are otherwise inaccessible by conventional gold panning . Subsequent distillation of the mercury leaves a crude residue of gold and silver . Unfortunately , mercury is quite volatile , even at room temperature , and is both physiologically toxic to humans and ecotoxic to other living organisms in the environment . Mercury pollution from the gold miners has become a major problem in the Amazon watershed and ecosystem .

The base metal lead component was absorbed into , and chemically combined with the bone ash or magnesium oxide of the cupel . Any noble metals present in the lead button remain unreacted on the upper concave surface of the cupel , which usually has the profile of a crucible (but is “solid”, of course) . Again in my college student days I had conversations with a mining student about his experiences with cupellation . As part of his laboratory training program he learned how to do cupellation analyses on gold and silver ores . My enterprising friend surreptitiously retained the small gold and silver beads he and his classmates produced in their ore analyses . He stored these precious metal droplets in a jar , and after graduation he sold them for a tidy sum of money to a gold dealer , helping to pay off his tuition fees ! This was , he laughingly told me , “high-grading” ; generally , in the mining industry high-grading is any stripping out of the valuable high grade ore , while ignoring or discarding the lower grade material .

One intriguing comment he made about cupellation stuck in my memory . Apparently with lower grade ores the droplets were rounded , with a smooth surface ; but with higher gold and silver concentrations the noble metal bead left on the cupel surface had a rough , jagged , appearance . In any case , I received a preliminary verbal report over the telephone from the geology firm in Ottawa about a week after sending the iron button to them . Three decades has dimmed my memory of the exact numerical data in this investigation , but I believe the platinum result was around 1200 ppm concentration in the iron button , which would translate into 120 ppm (i.e. one-tenth) in serpentine mill tailings . The analyst mentioned that the metal bead had a jagged appearance ; I could tell from his voice that he was very interested in the source of this strange metal button I had sent him , but I “played innocent”, pretending I was unaware of any significance of the result he had cited .

What was its significance ? Platinum is quite a rare element , occurring at 0.005 ppm in the Earth's crust . Infrequent high concentrations of it are found as placer deposits in widely dispersed locations such as in Russia and Colombia . Elsewhere , even its economically exploitable ores have only a few ppm Pt concentration. The Bushveld igneous complex (Merensky Reef) platinum ores in South Africa have a PGM concentration of 3-6 ppm (or 4-7 ppm) , which can be beneficiated to 100-1000 ppm by froth flotation . George states that the PGM concentration in the Stillwater complex in Montana is from 17.3-19.5 ppm , of which approximately 80% is palladium and 20% is platinum . Platinum concentration in nickel ore mined in the Sudbury , Ontario region is a mere 0.5 ppm . However , various by-product metals such as cobalt and the PGMs can be economically extracted during the nickel production , despite their very low concentration in the ore , thereby helping to make the overall mining operation profitable .

So serpentine with a Pt concentration of 120 ppm would be a fabulous platinum ore . And given the mountains literally of asbestos tailings that have accumulated over the years and decades in the Eastern Townships , and continue to accumulate on a daily basis , we could be looking at a PGM source of incredible value . Wealth beyond the wildest dreams of avarice ! In an excited and elated mood I began to draft a lengthy and detailed memorandum to management about this momentous discovery . Alas , if only I had heeded the cynical old comment , “If it looks too good to be true , it probably is”. And it was .


The Platinum Mine is Lost


I worked industriously late into the evening of that day to complete the report on the platinum analysis for our laboratory director (in those pre-computer days paperwork was drafted by hand before being completed on an electric typewriter ; this led , in my case , to severe hand cramps !) . I dropped the manuscript off on my pool secretary's desk on my way out of the Research Center , and back home . The next morning I lurched back into the laboratory , bleary-eyed after a sleepless night , to recommence my regularly scheduled research program the platinum had been an engaging diversion when a worried-looking researcher from the High Temperature group approached me with dismaying news . The temperature of the alkali fusion reaction had been monitored with an immersion thermocouple ; it was of the platinum-rhodium type , sheathed in an alumina jacket . And guess what ? They discovered after completion of the reaction and clean-up that it was missing ! Gosh , I wonder where it went ?

Of course , I knew right away where it had gone , and with it vanished the platinum mine . By way of explanation a little more chemistry discussion will be in order . It was obvious that the alumina sheath of the thermocouple had dissolved in the molten alkali . The industrial mineral bauxite is a very crude form of alumina , with substantial silica , titania , and iron oxide impurities . The Bayer process is used world-wide on an immense scale to refine bauxite into pure aluminum oxide . In the Hall-Héroult process the purified alumina is dissolved in molten cryolite (Na3AlF6) and electrolyzed to produce aluminum metal . The Bayer process takes advantage of the fact that alumina is an amphoteric compound , reacting readily with , and dissolving in both aqueous acid and alkali , the latter conditions being used in the Bayer process . The Wikipedia article about it comments :

“A few years earlier , Louis Le Chatelier in France developed a method for making alumina by heating bauxite in sodium carbonate , Na2CO3 , at 1200 °C , leaching the sodium aluminate formed with water , then precipitating aluminium hydroxide by carbon dioxide , CO2 , which was then filtered and dried . This process was abandoned in favor of the Bayer process”.

So there went the alumina sheath . As for the platinum-rhodium thermocouple , it too had dissolved in the molten alkali . Platinum and the PGMs generally are fairly unreactive , as noble metals , to most chemical reagents and at most temperatures . Thus , platinum is one of the very few elementary metals that occurs naturally as nuggets like gold in placer deposits . Platinum will , like gold , dissolve in aqua regia , a solution of concentrated nitric and hydrochloric acids . Noble metals will also dissolve in aqueous solutions containing strongly nucleophilic anions or other molecular species capable of forming strong coordinate covalent bonds with their cations . An oxidizing agent of some sort or other (commonly air , or pure oxygen , or hydrogen peroxide) must be added to the dissolving solution to oxidize the zerovalent noble metal atom to its corresponding cation .

A nice example of this is gold cyanidation , which is the principal method of extracting gold from its ores , or from gold dust in alluvial deposits (a better way than by the mercury extraction mentioned above , but still not very environmentally friendly , as cyanide pollution from flooded or breached leaching ponds has occurred a half dozen or so times around the world , causing considerable damage to the local watershed ecologies) . In 1887 the MacArthur-Forrest process was developed , in which an alkaline slurry of the crushed gold ore is vigorously stirred with a cyanide solution , with the simultaneous addition of an oxidixer usually air , pure oxygen , or H2O2 to convert the Au0 to Au1+, which then forms the very stable [NCAuCN]1- coordinate covalent complex :

2 Au0 + 4 (CN)- + ½ O2 (g) + H2O -------------> 2 [Au(CN)2]1- + 2 OH- .

The aqueous [Au(CN)2]1- solution is subsequently treated with a reducing agent (commonly the cheap zinc dust) to release the gold as a precipitated powder , which is filtered , washed , dried , and melted into a crude ingot form .

Very likely platinum and the other PGMs would behave similarly in the MacArthur-Forrest process . In 1959 Randall and Ward reported the reaction of PGMs with alkaline earth oxides in an oxidizing atmosphere in their synthesis of perovskites . This same sort of oxidation could have resulted in the solution of the platinum-rhodium thermocouple in the molten soda ash flux . Subsequently , the carbon in the reactant mixture would have reduced all of the Transition metal cations present to their respective zerovalent atoms , which dissolved in the iron globules and buttons at the bottom of the graphite crucible .

In any case , it was quite obvious that the platinum-rhodium thermocouple had also dissolved in the molten soda ash , and its noble metal content was accurately detected in the cupellation analysis of the iron button into which it had been extracted during the alkaline fusion . Although disappointed , our open-minded management nevertheless continued to support my original hypothesis regarding the occurrence of PGMs in the serpentine tailings . Another fusion experiment was organized , this time using an optical (i.e. non-contact) pyrometer for monitoring the reaction temperature . More iron buttons were prepared and were sent out for chemical analysis .

This time the results for the PGMs were much lower as anticipated but still very interesting . No cupellation analysis was reported ; I suspect that the Ottawa analyst may have carried out a cupellation fire assay on one of the new iron buttons , but found no measurable residue on it . Instead , the trace metal analysis reported to us was that of a very modern method , neutron activation , performed at McMaster University , Hamilton , Ontario . In this technique the iron button was bathed in a neutron flux inside a reactor . Highly unstable isotopes of the PGMs were formed and rapidly decayed back into stabler nuclei . The decay patterns of the PGM isotopes are well-known , and the “decay spectrum” of the iron button was compared with them . From this spectrum quantitative data concerning the elementary composition of the button could be calculated . All six of the PGMs can be detected at a concentration of less than one microgram (mg) in the sample :

This sketch was copied from the web page , “Centre for Neutron Activation Analysis” [of McMaster University] . My thanks to the copyright holder .

Neutron activation analysis of the iron buttons confirmed the presence of PGMs in them , but at a very low concentration . I hope the reader will forgive my somewhat hazy recollection of the numerical data after three decades , but I believe the figure was about 10 ppm of platinum in the button , or ~1 ppm in the serpentine tailings . The other five PGMs (Pd , Rh , Ru , Os , and Ir) were also found in the button , but in lesser amounts than the platinum ; maybe 5 ppm or so of palladium . The McMaster analyst also found about 5% of nickel and 0.5% cobalt in it . And yes , they even reported 0.5 ppm of rhenium in the button as sort of a bonus , although we hadn't asked for a rhenium analysis . As these analytical results were somewhat discouraging , the “platinum project” was terminated , and all prospects for a platinum mine in the Eastern Townships disappeared . Or have they ?




Although by no means definitive , the neutron activation results obtained back in the early 1980s for the iron buttons are proof of at least a low concentration of platinum in serpentine tailings . If the figure of 1 ppm (1 gram Pt per metric tonne of tailings) is reasonably accurate , it indicates a concentration similar to that of the PGMs in the nickel ore of the “Nickel Belt” in the Sudbury , Ontario region . There the PGMs are extracted as a profitable by-product of the nickel processing . Could the same approach be taken with PGMs in the serpentine mill waste of the Eastern Townships asbestos industry ? That would imply the large-scale utilization of the tailings in high temperature reductive fusion processes in which the iron buttons are generated as a valuable by-product , and are subsequently processed for their various metal components , including the PGMs .

The sodium magnesium silicate produced from the calcined tailings in our Research Center's alkali fusion process had little economic potential . It would have to compete with the well established sodium silicate (“water glass”) in its commercial and industrial applications . I suggested substituting potassium carbonate for the soda ash and so producing potassium magnesium silicate instead . Nitrogen , potassium , and phosphorus (generally referred to as NPK) are the three major plant nutrients ; magnesium and sulfur (the latter from sulfate) are two minor nutrients . The K/Mg ratio could be adjusted to provide a product that was slowly soluble in water , and could thus be used as a slow release potassium and magnesium fertilizer (and being mildly alkaline it would also be effective in acidic soils) . However , the alkaline flux and potassium source , K2CO3 , was thought to be too expensive , and therefore the resulting KMg silicate product would be commercially uneconomical , so the idea was abandoned .

The other high temperature fusion process which might generate both a useful major product as well as the iron buttons by-product is Albert Winer's mineral wool insulation , mentioned above . It's an attractive material , even more so these days with the current interest in saving energy by insulating houses and other buildings . While it needs a little more research work to make it water-resistant , I think it's a promising outlet for serpentine tailings that deserves a closer second look .

There was an exhibition of various industrial materials produced by the research groups in a display case in the lobby of our Research Center . One of the exhibits was a beautiful , shiny , hemispherical ingot of pure magnesium metal . It had been synthesized from calcined serpentine tailings (approximately Mg3Si2O7) by a variation of the Pidgeon process :

Mg3Si2O5(OH)4 ------ (heat , 2 H2O) ------> Mg3Si2O7 -------

(+ 1½ Si0 , high vacuum , heat) -------> 3 Mg0 (g) + 3½ SiO2 (c) .

In the actual Pidgeon process calcined dolomite (CaO.MgO) and ferrosilicon (roughly FeSi) are heated together in a furnace under a high vacuum , with distillation of the magnesium vapor into a cooled steel retort . Magnesium melts at 650 ºC and boils at 1107 ºC at one atmosphere pressure ; its boiling point is much lower in the very high vacuum used in the Pidgeon process . The remaining residue consists of a calcium iron silicate which can be used in the manufacture of Portland cement . Our Research Center magnesium group used an industrial grade of silicon as the reducing agent . I kept a chunk of it as a “souvenir”, which I still have here :

There are two interesting possibilities concerning this process . First , quicklime (CaO , from calcined limestone , CaCO3) could be combined with the calcined tailings and silicon to produce the magnesium metal vapor plus a calcium silicate slag , which could be utilized by a Portland cement manufacturer :

Mg3Si2O7 + 3½ CaO + 1½ Si0 ------- (high vacuum , heat) -------> 3 Mg0 (g) + 3½ CaSiO3 (c) .

In this case the feruginous components of the tailings (~ 10% by weight) , including their PGMs , would remain in the calcium silicate clinker ; no iron buttons would be produced in the reaction .

Second , the alkaline fusion reaction of the calcined tailings with soda ash could be carried out as usual ; this would produce the water-soluble type of sodium magnesium silicate together with the iron buttons . The former product would be leached with water to separate the soluble sodium silicate and the insoluble magnesium hydroxide , which would be filtered , dried , then dehydrated to magnesium oxide . This MgO would be reduced to magnesium by silicon , with the added CaO :

Mg3Si2O7 + 2 Na2CO3 ------- (fuse in an arc furnace) -------> Na4Mg3Si2O9 + 2 CO2 (g) ;

Na4Mg3Si2O9 + 3 H2O ---------> 2 Na2SiO3 + 3 Mg(OH)2 ----- (heat , 3 H2O) -----> 3 MgO ;

2 MgO + Si0 + CaO ------- (high vacuum , heat) -------> 2 Mg0 (g) + CaSiO3 (c) .

The four products of this overall process would therefore be magnesium metal , sodium silicate , calcium silicate , and the iron buttons . The buttons could be stockpiled and periodically shipped to a suitable metallurgical processor for further treatment . One possible procedure to extract the valuable components of the buttons might be the Mond process , in which they would be ground to a fine powder , then heated in a stream of pure carbon monoxide to form their respective Transition metal carbonyls . Iron carbonyl [pentacarbonyliron(0)] boils at 103 ºC , and nickel carbonyl [tetracarbonylnickel(0)] boils at 43 ºC , so the Fe and Ni could be cleanly separated . Cobalt carbonyl [octacarbonyldicobalt(0)] and chromium carbonyl [hexacarbonylchromium(0)] are solids , melting at 51 ºC and 154 ºC , respectively . While carbonyls of Rh , Ru , Ir , and Os are known (I'm not sure about those of Pt and Pd) , they have complex molecular structures and I doubt they would form in typical Mond process conditions . They would probably remain in their elementary form as residues in the Mond reactor , and could be separated and recovered by conventional chemical techniques .

Two criticisms of the overall process outlined above as applied to the serpentine-derived Mg3Si2O7 and MgO is that first , it would be very energy-intensive ; and second , it would be a batch process , which is economically less practical than a continuous flow process for magnesium production . As to the first concern , let's face it : any magnesium extraction method , whether electrochemical (as in the electrolysis of anhydrous MgCl2) or thermal reduction (Pidgeon process) is bound to be energy-intensive because of the relatively high E0red of Mg2+ (2.372 V) . Considerable energy is required to reduce Mg2+ to Mg0 . Fortunately , Quebec in which the vast serpentine resources are located is blessed with an abundance of inexpensive hydroelectricity , which continues to be developed and expanded . This cheap electrical energy is already used in the Quebec aluminum industry , one of the largest in the world . Magnesium metal production from serpentine could become another energy-intensive Quebec specialty product .

As for the second drawback , that of a supposedly uneconomical batch process method , China now dominates the world magnesium market , using the Pidgeon process with local supplies of dolomite and magnesite . Chinese producers are said to supply around 65% of the global production of magnesium metal at this time . I'm a chemist , not an economist , but I think that with abundant , cheap hydropower and huge deposits of serpentine Quebec could become a competitive producer of magnesium metal in the world marketplace .

And the platinum and other PGMs in the asbestos tailings are still there , just waiting for the right chemical process to make them economically available and attractive for exploitation . So my vision of an Eastern Townships platinum mine may still be realistic ; but as this story has shown , it should be tempered with cautious realism and supported by careful scientific research , hard work , and good chemistry .


References and Notes


considerable interest : The Magnola plant was built in 1998-2000 near the town of Danville , in the Eastern Townships , to extract magnesium from asbestos tailings . They were leached with hydrochloric acid to produce a MgCl2 brine , which was fully dehydrated to anhydrous MgCl2 . This salt was melted and electrolyzed to magnesium metal (cathode) and chlorine gas (anode) . The chlorine was combined with hydrogen to produce HCl , which was recycled back into the extraction process . A detailed description of the extraction chemistry and geochemistry is provided by :

T. T. Chen , J. E. Dutrizac , and C. White , “Serpentine Ore Microtextures Occurring in the Magnola Magnesium Process”, J. Miner. Met. Mater. Soc. 52 (4) , pp. 20-22 (2000) .

The Magnola plant cost about $730,000,000 to build , and was designed to produce ~ 60,000 tonnes/yr of magnesium metal ; it was in operation for only a few years and then was shut down , apparently because it couldn't compete economically with cheaper Chinese magnesium (from the Pidgeon process ; see above near the end of the text) . Finally , it was completely torn down , with no buildings now remaining on the former plant site .

Why was this project such a complete failure ? One hint is provided by this web page , Magnola Health Effects . It seems that the MgCl2 electrolysis was the source of worrisome emissions of hexachlorobenzene (HCB) and dioxins (TCDD) :

Where did they come from ? I believe that the anodes in the electrolyzers , at which the chlorine gas by-product was produced , must have been made from carbon , probably graphite . The hot , nascent chlorine atoms are extremely reactive , and they must have torn chunks of aromatic rings off the graphite surface , combining with them to form the HCB and TCDD . I'm surprised that the chemical engineers designing and operating the electrolyzers couldn't have found a more unreactive anode material than graphite . After all , the Dow Chemical Co. had been successfully using a similar MgCl2 electrolysis process for almost six decades in their Freeport ,Texas magnesium plant and presumably didn't observe such unwelcome toxic emissions (P.D.V. Manning , “Magnesium , Metal of the Future”, Engineering and Science Monthly , pp. 14-18 , June , 1944 , PDF , 860 KB) . The electrolysis of molten chlorides is a mature technology , holding few if any unpleasant surprises for the unwary operators . The Downs cell , in which sodium metal is produced from sodium chloride , and lithium metal is derived from lithium chloride , is a good example of an efficient electrolyzer for molten chloride salts . Chlorine gas is produced at the anodes which are generally made of graphite in the Downs cell .

Another possible problem for the Magnola plant may have been in the magnesium metal product . I heard a rumor to the effect that it might have contained a small amount of nickel impurity , which slightly degraded its metallurgical specifications . The nickel , of course , would have originated in the feruginous phase of the serpentine tailings feedstock material , as discussed above . Aqueous acid leaching certainly would have extracted some of the Transition metal components of the tailings along with the magnesium :

“Magnesium leaches rapidly from the serpentine Mg3Si2O5(OH)4 , and the silicon remains in-situ as an amorphous silica pseudomorph after the original serpentine particles . Negligible silica dissolution occurs , and silica gelation was never observed . The reaction interface extends over 300–400 µm ; as a consequence , fine grinding does not significantly accelerate the rate of magnesium dissolution . Associated inclusions of brucite Mg(OH)2 , awaruite Ni8Fe3 , and magnetite Fe3O4 dissolve rapidly ; whereas , chromite FeCr2O4 and a chromium-rich spinel (Cr,Fe,Al,Mg)3O4 remain largely unaffected”.

T. T. Chen , J. E. Dutrizac , and C. White , op. cit. (from their abstract) .

The magnesium metal from the Pidgeon process , on the other hand , is completely free of nickel or other Transition metal impurities , as they remain as non-volatile residue (clinker) in the distillation pot of the still , while the ultrapure magnesium vapor distils into the cooled reception vessel of the retort . Magnesium from the Pidgeon process typically has a purity of up to 99.99% Mg ; that from MgCl2 electrolysis is of lesser purity , generally 99.8% Mg . Thus , the competing Chinese magnesium doesn't have any nickel contamination and meets all the required metallurgical specifications . Also , if the Transition metal content of the tailings was removed from them as iron buttons in an alkali fusion process , the resulting MgO feedstock would be even purer than the tailings with respect to any nickel impurities .

A third general sort of chemical problem with the Magnola process was its acidic chemistry , utilizing the extremely corrosive hydrochloric acid , and with the very corrosive , reactive chlorine by-product . As a former practicing chemist I've used hydrochloric acid in various strengths , hydrogen chloride gas , and chlorine , and all three are difficult reagents to handle and work with . These corrosive , acidic , reactive chemicals must have put quite a strain on the Magnola plant infrastructure . An alkaline fusion process such as I described above , together with the chemically neutral Pidgeon process , would be much less demanding from a corrosion point of view .

Summarizing , I believe three technical problems (apart from the publicly stated economic one about the cheaper Chinese magnesium) may have contributed to the failure of the Magnola plant :

unanticipated toxic organic emissions (HCB , TCDD) from the MgCl2 electrolyzers ;

possible nickel impurities might have degraded the magnesium's metallurgical properties ; and ,

corrosive HCl (aq) , HCl (g) , and Cl2 (g) severely stressed the plant infrastructure .

All three problems would be avoided by an alkaline fusion method , as was described above ; in addition , the PGMs , Ni , Co , Cr , and Fe could be recovered as valuable by-products , helping to make the overall extraction process economically feasible . The Magnola magnesium plant is an excellent example of the right idea , but the WRONG chemistry.

The magnesium industry apparently is a highly competitive , even cut-throat business ; Dow Chemical Co. finally ceased magnesium production in 1998 due to adverse production problems and intense pressure from cheap imported magnesium . For an interesting and quite readable economic analysis of Dow's magnesium history , see : M. B. Lieberman , “The Magnesium Industry in Transition”, 11 pp. , August , 2000 [PDF , 74 KB] .

Albanian : S. Shallari , C. Schwartz , A. Hasko , and J. L. Morel , “Heavy Metals in Soils and Plants of Serpentine and Industrial Sites of Albania”, Sci. Total Environ. 209 (2-3) , pp. 133-142 (1998) .

Spanish : A. Paz-Gonzalez , M. T. Taboada Castro , and S. R. Vieira , “Geostatistical Analysis of Heavy Metals in a One-Hectare Plot Under Natural Vegetation in a Serpentine Area”, Can. J. Soil Sci. 81 (3) , pp. 469-479 (2001) [PDF , 702 KB] .

Albert Winer : A report of Winer's mineral wool has been published in : Canadian Mining & Metallurgical Bulletin 67 , pp. 97-104 (1974) , which I've been unable to obtain a copy of . Mr. Winer mentioned to me that he learned the technique of spinning his mineral wool from experimenting with a cotton candy machine , such as you might see at a county fair or carnival ! Mineral wool manufacture is described in Wikipedia , Eurima , and in an EPA document [PDF , 39 KB] .

merely heavier members : Of course , by this same reasoning PGMs should also occur to a significant extent in iron ore (hematite , magnetite , limonite , siderite , goethite) . Do they ? PGMs actually do occur in South African chromite and in nickel minerals in the Sudbury , Ontario region , so they could probably be detected in iron ore as well . I think most of the PGM elements are actually located in the molten iron core of the Earth , which has acted as a sort of planetary-sized cupellation bead, dissolving them during the formation of Earth millions of years ago . Only a relatively small percentage of the heavy metal elements , including the PGMs , remain in the outer silaceous crust of the Earth , where they are accessible to human miners .

I recall reading a series of adventure stories for children about an American whiz-kid inventor , Tom Swift , when I was a teenager (half a century ago !) . One of these books I enjoyed then was called Tom Swift and His Atomic Earth Blaster , written around 1954 by James Duncan Lawrence . The always intrepid and resourceful Tom Swift , who was evidently endowed with an enormous research and development budget , invented a nuclear-powered mining drill that was able to penetrate to the Earth's core and tap a huge spout of molten metal (iron-nickel) from it . Wonderful , imaginative science fiction !

sold them for a tidy sum : I'm reminded by my friend's cupellation bead high-grading of another such story of keen observation and opportunistic entrepreneurship . Unfortunately it must remain undocumented here , as I've been unable to locate the original reference in which I read about it . In any case it concerns a worker or supervisor in a metallurgical refinery who noticed that the metal ingots possibly of lead , I don't recall in the storage shed were sweating or weeping shiny metal droplets . Others had noticed them , too , but had ignored them . This particular person , though , knew what they were , and asked permission from his management to collect them , on his own time , of course . Since there was no interest in the metal droplets by his company , he was given permission , and it became his hobby to scrape them from the stacks of ingots in the storage shed . He continued collecting the metal beads for quite some time ; at his retirement from the company he had accumulated a large amount of the semi-liquid metal , which he sold somewhere for a substantial amount of money : a nice nest-egg for his retirement years ! He had recognized the metal droplets as gallium , which melts at 29.8 ºC , a few degrees above room temperature . Gallium is a rare , costly element having numerous applications in the semiconductor industry , for example in electronic devices using gallium arsenide , GaAs .

George : M.W. George , Platinum-Group Metals”, USGS Minerals Yearbook , pp. 57.1-57.12 (2004) [PDF , 490 KB] .

would behave similarly : Platinum(II) forms a square planar coordinate covalent complex with cyanide anions . The water-soluble salt-like compound K2Pt(CN)4 is well-known . It can be partially oxidized by bromine to form KCP, K2Pt(CN)4Br0.3 . 3H2O . This mixed-valent platinum compound is metallic , having the form of slender , lustrous , copper-colored needles with an appreciable electrical conductivity . I've written about KCP in several Chemexplore web pages , eg. New Solar Cells from Mixed-Valent Metallic Compounds”.

Randall and Ward : J.J. Randall and R. Ward , The Preparation of Some Ternary Oxides of the Platinum Metals”, J. Amer. Chem. Soc. 81 (11) , pp. 2629–2631 (1959) ; R. J. Bouchard and J. F. Weiher , “ LaxSr1-xRuO3 : A New Perovskite Series”, J. Solid State Chem. 4 (1) , pp. 80-86 (1972) .

open-minded management : The director of our Research Center in the early 1980s was Dr. Jean-Marc Lalancette , who previously was a professor of chemistry at the nearby Université de Sherbrooke , Quebec . Dr. Lalancette was a creative , inventive , “chemicals chemist” ; one of the commercial products he devised was “Seloxcette” (chromium trioxide intercalated in graphite) . This reagent is a mild , selective oxidizing agent which is useful in organic chemistry syntheses :

J.-M. Lalancette , G. Rollin , and P. Dumas , “Metals Intercalated in Graphite . I . Reduction and Oxidation”, Can. J. Chem. 50 (18) , pp. 3058-3062 (1972) [PDF , 259 KB] ; J.-M. Lalancette , “Oxidation of Primary Alcohols”, U.S. Patent 3840566 , October 8 , 1974 .

Another very noteworthy graphite intercalation compound he prepared is that of antimony pentafluoride in graphite . Presumably the intention was to use this material as a mild Friedel-Crafts catalyst , in place of the very toxic and hazardous liquid reagent SbF5 :

J.-M. Lalancette and J. Lafontaine , “Intercalation of Antimony Pentafluoride in the Lattice of Graphite”, J.C.S. Chem. Commun. 1973 , p. 815 ; J.-M. Lalancette , “Graphite Intercalated Antimony Pentafluoride”, U.S. Patent 3950262 , April 13 , 1976 .

I've long been fascinated by metallic solids and electrically-conducting materials . Dr. Lalancette's SbF5–graphite intercalation compound (75% w/w) was of particular interest to me because it has the highest known ambient electrical conductivity of any known material (obviously , the superconductors have a higher conductivity , but not at room temperature !) . Several years after its initial preparation it was studied by an American researcher , F.L. Vogel , who found it had the incredible ambient electrical conductivity of ~ 1,000,000 ohm-1-cm-1 :

F.L. Vogel , “The Electrical Conductivity of Graphite Intercalated with Superacid Fluorides : Experiments with Antimony Pentafluoride”, J. Mater. Sci. 12 (5) , pp. 982-986 (1977) ; idem , “Intercalation Compounds of Graphite”, pp. 261-279 in W.E. Hatfield (ed.) , Molecular Metals , Plenum Press , New York , 1979 ; see especially Table III , “Acceptor Intercalated Compounds of Graphite”, p. 269 ; idem , “Process for Conducting Electricity Utilizing a Specifically Defined Graphite Intercalation Compound ”, U.S. Patent 4293450 , October 6 , 1981 .

By comparison , the second best electrical conductor is pure silver , far behind at 630,120 ohm-1-cm-1 . To the best of my knowledge , SbF5–graphite (75%) still holds the world record for the material with the highest ambient electrical conductivity . That's a nice achievement ! I've long admired Dr. Lalancette for his breadth of chemical knowledge , creativity , and perceptive intellect .

Pidgeon process : Wikipedia article ; article .

quicklime : A limestone quarry is still in operation at Lime Ridge , here in the Eastern Townships . Presumably most of its limestone product is destined for the manufacture of Portland cement , with a smaller amount diverted into quicklime , CaO , and slaked lime , Ca(OH)2 , for use in construction and agriculture (as a soil sweetener , i.e. to reduce acidity) .

batch process : The Kroll process is a good example of a successful batch process for the production of an industrial metal . Titanium tetrachloride is reduced to titanium by magnesium metal in a stainless steel pressure-cooker reactor :

TiCl4 (g) + 2 Mg0 (l) ------ heat [800 850 ºC] -----> Ti0 + 2 MgCl2 .

Incidently , this latter reaction might form the basis of another Quebec-based metallurgical industry : using magnesium metal from Townships asbestos tailings to manufacture titanium metal from ilmenite ore , which is mined in large quantities in northeastern Quebec . Ilmenite (FeTiO3) is chemically converted into purified titanium dioxide , mostly used as a brilliant , opaque , white paint pigment . Titanium dioxide can be reductively chlorinated to titanium tetrachloride :

TiO2 + C + 2 Cl2 (g) -----------> TiCl4 + CO2 (g) .

As mentioned above , Quebec already has one of the largest aluminum industries in the world , due mostly to its inexpensive hydroelectricity . It also has the mineral and energy resources to be a world leader in the production of two more important light metals , magnesium and titanium , which are of key importance in the automobile and aerospace industries . Of the two I consider titanium to be the far more valuable one . With the lightness of aluminum , the strength of steel , a melting point greater than that of iron , high chemical resistance (including non-rusting) , and an incredible toughness , titanium is truly a super-metal. I think it has a great potential in future high-tech vehicle , aircraft , spacecraft , and construction applications . Magnesium could be produced as a sort of captive product, used internally in Quebec only to manufacture titanium via the Kroll process , since it might not be able to compete globally with the inexpensive Chinese magnesium .

I had an idea about titanium : might it be possible to dissolve titanium dioxide in a molten fluoride salt (or a eutectic mixture of ionic salts , such as LiF–NaF–KF) and electrolyze it , in an analogy of the Hall-Héroult process for aluminum mentioned above ? A cathode of pure titanium , on which the electrolytic titanium would be deposited , would have to be slowly raised as the electrolysis proceeded (titanium melts at 1668 ºC , which is much higher than the temperature of the molten salt bath would be) , and periodically cut off or replaced as it became too large . The oxide anions would be oxidized to oxygen atoms at the carbon anode , which would be gradually burned to CO2 , as in the Hall-Héroult process . If it was successful , this TiO2 electrolysis could replace the Kroll process and permit the large-scale production of very pure titanium metal . It would be an excellent research project !



[ Index Page ] [ Contact ]