In The Graduate (1967), Mr. McGuire told Benjamin, “There’s a great future in plastics.” Those “plastics” were a metaphor for the phoniness of (adult) society, but as a chemist, I have to admit, McGuire was right. And there’s an even greater future for plastics (and other polymers) in the 1632verse.
I have divided this article into two parts. In part A, after a brief overview of polymers, I present my proposed timeline for polymer development, and then discuss, in the expected chronological order of availability, the polymers and their properties and potential uses. I will also cover industrial polymerization methods, methods of forming the polymers into useful articles, polymer blends and additives, composites, and (incompletely) polymer economics.
Some readers will no doubt be troubled by the placement of certain very important polymers late in the new timeline. In part B, I go into some detail regarding the problems confronting the polymer industry, both in general and with particular regard to my “late” polymers.
PART A: POLYMERS: USES AND PREDICTED AVAILABILITY
Use-Types of Polymers
Functionally, polymers include plastics, fibers, elastomers (rubbers), adhesives and surface coatings. 73% of synthetic polymer production in 2000 was for plastics, 13% for fibers, 9% for coatings and adhesives and 5% for rubbers. (Elias 11).
Plastics are dimensionally stable at room temperature but at least initially can deform (change shape) under the influence of heat and pressure. The deformation is inelastic, that is, they don’t spring back to its original shape after it cools. Elastomers stretch under stress at room temperature, and recover their shape when relaxed. Fibers have lengths great in proportion to their diameters, and high tensile strength. They can be woven into fabrics, or introduced into a matrix made of plastic or other materials to form a fiber-reinforced composite. Coatings adhere to a surface to decorate or protect it, and must be resistant to abrasion. Adhesives must adhere to two, possibly chemically different, surfaces, in order to join them. They must at the same time be sufficiently cohesive so as to hold the joined surfaces together.
Plastics and elastomers are mutually exclusive categories. (Although light crosslinking may convert a plastic to a rubber.) However, there are polymers we think of as fibers (nylon) that can be used to make bulk plastic parts, and ones we think of as bulk plastics (polyethylene, polypropylene) that can be drawn to make useful fibers. Plastics and rubbers can also be compounded into formulations suitable for adhesive or coating uses.
Plastics Use in the Twentieth Century
Plastics (resins) flow under pressure and therefore can be formed into a desired shape. As a result, even if the polymer is expensive, the finished article might not be, because the fabrication costs are low. Generally speaking, plastics are low in density and in electrical and heat conductivity, and chemically resistant.
Historically, synthetic plastics were developed as substitutes for ivory, tortoiseshell and horn. Plastics can also serve as substitutes for:
glass (transparent plastics are lighter and can have greater impact resistance)
paper and leather (plastics can be less permeable to fluids)
wood (plastics have greater dimensional stability and resistance to biodegradation)
metal (plastics can have an excellent strength/weight ratio and provide low-friction surfaces).
About half of modern plastics use falls into the following categories, with the most popular plastics (as of 1988) indicated in parentheses:
packaging—containers, foam, films, coatings, etc. (high density polyethylene (HDPE) or low density polyethylene (LDPE), polypropylene (PP), polyethylene (PE) or polybutylene terephthalate (PET), polyvinyl chloride (PVC), polystyrene (PS))
building and construction—plumbing, resin-bonded wood composites, flooring, siding, heat insulation, windows, etc. (PVC, phenolic, HDPE, urea-formaldehyde, amino resins, LDPE, polystyrene, polymethyl methacrylate, acrylonitrile-butadiene-styrene (ABS))
electrical/electronic—wire insulation, housings, plugs and sockets, printed circuit boards, fiber optics (PVC, polystyrene, LDPE, HDPE, ABS, phenolic resins, polyamides, epoxides, polycarbonates)
furniture—structural components, padding, wheels and casters, coverings (phenolic, polystyrene, PVC, polypropylene).
household goods—small appliances, toys, sporting goods (LDPE, HDPE, polystyrene, polypropylene)
transportation—vehicle bodies, instrument panels, tires and wheels, adhesives (polypropylene, PVC, ABS, polyamides, urea-formaldehyde, epoxides)
(Elias 9; Witcoff II-9ff, Muccio 5ff).
Introduction to the Polymer Timeline
The dates that various polymers were discovered and commercialized (often much later) in the old time line are available (Ram 2), but it’s clear that the polymer industry will develop quite differently in the new time line.
My conclusions as to when experimental quantities of the monomer (enough to start experimenting with polymerization processes) could plausibly be available are set forth in Table 5-1. This assumes a fairly smooth development process and it could easily take longer than what I have proposed.
Even so, I must consider early (1630s) production of synthetic polyethylene (or polymethylene), polypropylene, poly-cis-butadiene, poly-cis-isoprene, or their copolymers to be very much a long shot. (Don’t believe me? Read part B . . . .)
For any of the polymers, figure that it could take another 5–20 years to scale up production of the monomer(s) and the polymer to the point that the compound is readily available.
Table 5-1: Optimistic Polymer Availability Timeline
cellulosic fibers such as cotton, silk; shellac, casein and keratin; some native use of natural rubber, balata, gutta-percha, chicle.
coagulated casein plastic*
November 1633: natural rubber from Hevea brasiliensis*; December 1633: regenerated cellulose (viscose)*
natural rubber from Hevea guianensis*, Castilla elastica*, milkweed*, balata*, vulcanized natural rubber*
casein-formaldehyde plastic, urea-formaldehyde and phenol-formaldehyde (1634*) resins, polyoxymethylene (polyacetal) resin, cellulose nitrate, cellulose acetate, rayon (1633*), polyglycolic acid, polylactic acid, furan resin
melamine-formaldehyde resin; polystyrene, semisynthetic impact polystyrene; polyurethane, polyethylene terephthalate, Dacron®, polycarbonate, polyphenylene sulfide, polyvinyl chloride and related polymers, polyethylene oxide (glycol), polypropylene oxide (glycol); butyl rubber; gutta-percha, chicle
polyphenylene oxide; epoxy resin; Thiokol rubber; neoprene (polychloroprene), polyhydroxybutyrate; natural fumaric acid-based unsaturated polyesters?
nylon 6,6; polycaprolactone, nylon 6; aramid; polyimide; polymethyl methacrylate, polymethyl cyanoacrylate; polyacrylonitrile; styrene-acrylonitrile rubber; polytetrafluorethylene (PFTE); polyvinyl fluoride; silicone rubber; acrylonitrile-isoprene-styrene (AIS) terpolymer? vinyl ester resins?
fully synthetic impact PS; styrene-butadiene rubber; ABS terpolymer; nitrile rubber; unsaturated polyester; styrene-maleic anhydride copolymer
pressure tech dependent: LDPE
catalyst tech dependent: polybutadiene, polyisoprene, HDPE, linear low density polyethylene (LLDPE), isotactic and syndiotactic PP; EPDM rubber
multiple issues: methyl rubber
In 2000, the leading plastics by world production volume were polyethylenes (52 million tons), polypropylenes (29 million tons), polystyrenes (26 million tons), polyvinyl chlorides (13 million tons), thermoplastic polyesters (10.5 million tons), and polyurethanes (8.4 million tons). (Elias 10). Clearly, in the first decade of the new time line, the leading plastics will be somewhat different.
Let’s look at the new time line’s first polymers in more detail.
Vulcanized Rubber. Natural rubber contains cis-polyisoprenes which make it elastic. See Cooper, “Bouncing Back: Bringing Rubber to Grantville” (Grantville Gazette 6). Heating masticated rubber with molten sulfur, or dissolving it in a solution of sulfur chloride in carbon disulfide causes crosslinking (vulcanization) of the polymer chains. Vulcanized rubber first appears in canon in winter 1633–34, see Offord, “Letters from France” (Grantville Gazette 12).
Light crosslinking (10 parts sulfur:100 parts rubber) results in a product that retains elasticity over a broader range of temperatures . . . even those of American winters and summers. Heavy crosslinking (50 parts sulfur) results in a plastic, called ebonite, vulcanite or hard rubber.
In response to recent queries on Baen’s Bar as to how much natural rubber is available in Europe in the 1634–35 period, in canon we have the following rubber sources developed:
1) Henrique da Costa organized a Portuguese Hevea brasiliensis tapping operation in Brazil, on the Tapajos. (Cooper, “Stretching Out, part 2: Amazon Adventure,” Grantville Gazette 12). Each tapper walked one route, of 50–100 trees, each day. The rubber trees were tapped every other day and each produced about 2 pounds rubber/collection month. The collection season began June 1633 and would have continued at least through November 1633 and possibly as late as February 1634 (whenever the waters have risen too high). So we are talking 12–18 pounds/season/tree. (Some sources say 10 pounds/season/tree.)
If we assume 50–100 tappers as a starting point, then the first season we would have produced 30,000–180,000 pounds (15–60 tons) of rubber. The rubber was carried by sugar ships stopping at Belem every month or so. It would have taken them at least two months to reach Lisbon. The Portuguese would eventually sell most of the rubber in the USE, since that’s where most of the demand would be. Henrique was forced to flee into the rainforest to avoid arrest, and we don’t know how soon there will be a second production season, under new management.
2) David de Vries and Maria Vorst established a bauxite/rubber-based colony in Suriname (Cooper, “Stretching Out, Part 3: Maria’s Mission,” Grantville Gazette 14). Their ships left Hamburg in December 1633 (presumably before any of the Brazilian rubber reached Hamburg) and arrived off the coast of Suriname in Early 1634. Upriver, they begin tapping Hevea guianensis. Surprisingly little is known about this tree, but I am going to assume it is essentially identical to H. brasiliensis. They definitely do not have as many tappers as Henrique, perhaps 20. Suriname, near the coast, has two dry and two rainy seasons, the dry seasons are February through April and mid-August through November. So 6–7 months collecting.
De Vries got stir-crazy and by April 1634 he was in Trinidad. It follows that he couldn’t be carrying much H. guianensis rubber on his ship, they weren’t tapping long enough. If he has the March production, that’s perhaps one ton of rubber. (In the 1634–35 season, perhaps they can produce 6–7 tons. More if they can recruit more tappers.)
3) Just to complicate matters, De Vries harvested Castilla elastica rubber in coastal Nicaragua in May-June 1634. Cooper, “Stretching Out, Part 4: Beyond the Line,” Grantville Gazette 16). There, they are cutting down, rather than tapping, the trees, each felled tree producing 15–20 pounds of rubber. (If they chose to tap and come back, note that Castilla can be tapped only 1–4 times/year and that one tap is good for no more than 1 pound, probably less.)
It’s difficult to guess how many trees they’ve found and cut down. Density estimates are anywhere from 50–700 trees/acre. Let’s say 1000 trees, producing 7.5–15 tons rubber. The rubber arrives in Hamburg in late 1634. Treat this as a one-time deal. (The trees take 6 years to mature, and the Spanish will increase coastal defenses since come summer De Vries led raiders inland, ransacking Granada’s gold and silver.)
4) Finally, I have seen reference to milkweed rubber being available in 1634. Offord, “Feng Shui for the Soul,” (Grantville Gazette 17). Milkweed is a bit problematic, because the rubber yield is low and there’s a high resin content. The yield per acre per year is 114–171 pounds. If we had 10 acres under cultivation in 1634, then that would mean producing half to three-quarters ton rubber.
Balata (50% trans-polyisoprene) comes from Manilkaea bidentata, found in Guyana and the West Indies. It has been used for machine belting. (EB15/balata). The colonists in my Suriname colony are collecting it, beginning in early 1634.
Casein and Casein-Formaldehyde Plastics. Primitive casein plastics have appeared in canon, as the result of treating milk with vinegar (dilute acetic acid). Offord, “Bootstrapping” (Grantville Gazette 11); DeMarce, “Songs and Ballads” (Grantville Gazette 14).
This is simply the commercial embodiment of a classic elementary school science experiment for making “plastic milk.” There isn’t really a chemical reaction, but rather the acetic acid is coagulating the casein (essentially the same mechanism as lactic acid curdling milk). The acid denatures (unfolds) the protein molecules, allowing them to aggregate. The casein is least soluble at its isolectric point, pH 4.6 (acidic). In OTL, the material is known as acid-precipitated casein and is used primarily in paper-coatings, but also as a paint emulsifier, adhesive, rubber additive (internal lubricant) and a fiber. (SIF)
A related material (rennet-precipitated casein) actually predates the Ring of Fire (RoF). A sixteenth century recipe says to extract a “paste” (crude casein) by repeatedly treating goat cheese with hot water and removing the liquid on top; then treat the paste with a heated soapy solution (alkali salts of fatty acids) and press it into a mold. The material, once cooled, imitates horn. The recipe was published by the Bavarian Benedictine monk Wolfgang Seidel (1492–1562), who had learned about it around 1530 from the merchant Bartholmaus Schobinger (1500–1585), who in turn had heard it from an alchemist. (Haefele 127; Elias 4; Deutsches-Kunstoff Museum). Presumably the soapy solution isn’t so alkaline as to convert the casein to soluble caseinates.
Rogers (837) recommended the addition of strontium hydroxide and powdered marble to increase the hardness and durability of casein plastic, and I believe that was what was done by Lilienthal when he used it to make children’s building blocks. (Elias 4, Scherer 98).
Rennet casein is reportedly still used to make buttons and other small plastic articles (SIF), and I have seen current advertisements for biodegradable casein knitting needles (arnos), but I strongly suspect that many of these products are not simple casein, but rather casein “hardened” with formaldehyde as I discuss below. For one thing, a primitive casein plastic will “dissolve” if it comes in contact with weak alkali. (Johnston, 969). Also, it is vulnerable to moisture attack ( Brady 182), and it’s brittle (Seidel said “bending or twisting will cause it to shatter like glass”).
Casein is a milk protein and thus already a polymer, but it can be further crosslinked by formaldehyde. The new polymer, the conventional “casein plastic” (galalith), was first produced in 1897. Either curds or whey may be used as starting material. Galalith is still used in the button industry. (Brydson 5). It has a tensile strength of 8000 psi, and can withstand a temperature of 150oC.
Not that casein-formaldehyde plastic production is without its trials. The dry process involved applying heat and pressure to casein powder using a screw press, thereby extruding a soft plastic, and then hardening it by immersion in formaldehyde solution. This hardening process took three weeks to six months, depending on the thickness of the material. Then the material had to be dried down to 8–12% moisture content, which took an equally long period. (Brothers).
While casein plastics can definitely be extruded, my sources (Brady 182; Salter, SIF, Wikipedia/Galalith) are hopelessly contradictory as to whether they can be molded.
For those interested in economics, I was able to find some information about the German experience with galalith in 1908. In France, they paid butter makers 30 cents/220 pounds of skim milk, coagulated it, and shipped the pressed curds back home for finishing. It took 2000 liters of skim milk to make 220 pounds dry casein, and that much casein was worth $15.50. A plant large enough to make one ton galalith/day would cost $300,000, and the factory in Hamburg, with 200 workers, produced 800 tons/year. (MCTR).
Cellulosics. Bracannot treated wood with nitric acid in 1833, creating cellulose nitrate (guncotton). Others similarly nitrated cellulose from paper and cotton. As of September 1633, Ferrara hadn’t yet made guncotton, but he knew how—the supply of nitric acid was still too limited to make cellulose nitrate in militarily significant quantities. (Weber and Flint, 1633, Chapter 28).
Historically, the immediate problem with cellulose nitrate was that it couldn’t be formed into a useful shape. (Deanin 167). In the 1870s, cellulose nitrate was plasticized with camphor, producing “celluloid”, the first thermoplastic. This was highly flammable, and only the degree of nitration distinguished it from guncotton explosive. In manufacturing facilities, “minor fires were frequent, major ones not rare,” and there was an explosion at one factory in 1888. (Meikle 22). Fabrication, storage and use were also dangerous. Workers sawed blanks of cellulose nitrate while working, stripped to the waist, under a stream of water. (79). Poor nitration could also be a problem, leading to the failure of Merchants Manufacturing Company. (19).
Camphor was a bottleneck in the production process, as it had to be imported from China, Japan or Formosa. Synthetic camphor was first made in Germany in 1901 and in the United States in 1933. (19). Also, a variety of new plasticizers were identified in the 1890s and thereafter. Plasticizers were required for other plastics, including cellulose acetate, butyrate, and propionate, and polyvinyl chloride, butyral and acetate, and by 1940 dioctyl phthalate was the leading general-purpose plasticizer. (Deanin 168).
Celluloid was used as a substitute for ivory, tortoiseshell and horn, as the middle layer in the first safety glass, and as the first movie film substrate, but its only significant modern use is in ping-pong balls.
Instead of nitrating cellulose, one can acetylate it, obtaining cellulose diacetate or triacetate. Similarly one can make cellulose propionate or cellulose acetate butyrate by reaction with the appropriate carboxylic acids.
Cellulose acetate was first synthesized in 1865 but not commercialized until the twentieth century (Meikle 78). It was first used as a “dope” for WW I airplane wings, and as a fiber. In the 1920s, it replaced celluloid for motion picture film, and it ultimately replaced celluloid in safety glass.
Cellulose acetate is potentially of great importance, because—if the plasticizer problem is solved—it will be the new time line’s first nonflammable thermoplastic. The development of injection molding encouraged its use as a strong, stiff, hard, tough plastic in, for example, protective goggles. Cellulose acetate butyrate and cellulose propionate aren’t as well documented in Grantville, but they’re superior in some respects to cellulose acetate.
Rayon (cellulose xanthate) was a “poor man’s silk,” developed in 1884, and given the huge prices commanded by Oriental silks, it will be an attractive target for the USE chemical industry. It appears that only “a limited quantity” of rayon is produced in early 1635 (Huff and Goodlett, “In the Army Now,” Grantville Gazette 20), but it was still more expensive than cotton, let alone silk. But I have discussed this with the authors, and they say that what they meant was that there are lots of uses for viscose, and the demand for it won’t be close to being filled as of 1635. So this isn’t inconsistent with rayon being available in June 1633 (as in their story “The Monster”) or even earlier.
The three basic methods of producing rayon are described in EB11/Cellulose and EA/Rayon. EB11 predicted that the viscose method would prevail, because of its lower cost, and EA says that it’s the most widely used. It requires sodium hydroxide (lye), already known to the down-timers, and carbon disulfide, in order to produce viscose (liquefied wood pulp cellulose). The viscose is extruded through spinerettes (holes) to make the fibers. If instead it’s extruded through slots, we obtain regenerated cellulose sheets that can be waterproofed to make cellophane.
If for some reason carbon disulfide is not available, the rayon can be made by treating cellulose with a copper-ammonium solution or with acetic and sulfuric acids. Acetate rayon, because of its high tensile strength, would be of interest to the fledgling airship industry. See Evans, “Wingless Wonders” (Grantville Gazette 19).
Phenolics. Phenol-formaldehyde (Bakelite) is a member of a broader class of plastics, the phenolic resins, made by reacting an aromatic alcohol (phenol, cresol, xylenol, bisphenol, resorcinol, etc.) with an aldehyde (formaldehyde, fufural). Furan resin (phenol-furfural) is another member of this class. There are also some ion-exchange resins (OTL 1935) that are related to Bakelite. (EB15/ion-exchange reaction).
Bakelite was used for insulation of coils, knobs and sockets for electrical equipment, adhesives for making plywood and particle board, and, with reinforcement, heat-resistant electrical connectors and appliance handles.
Furan resins are chemical and heat resistant. They’ve been used in pavement coating, chemical plant lining; and with asbestos fiber reinforcement, construction material. (EB15/heterocyclic compound; CCD401).
Urea-Formaldehyde resin (OTL 1925) had uses similar to phenol-formaldehyde, but offered a broader color range. (Meikle 76). The price of a grocery scale was reduced from 49 to 35 cents/pound by replacing the cast iron with urea-formaldehyde. (110).
Polyacetal. Engineering thermoplastic, with high chemical resistance and dimensional stability, good temperature and abrasion resistance, and low friction. In the old time line, polyacetal was commercialized in 1960 (Salamone 2). In the new time line, it’s likely to be the first engineering plastic. It’s useful as a replacement for small metal parts in appliances and machinery.
Polyglycolic and polylactic acid (PLA): Biodegradable. PLA was first made in the 1890s. (GreenPlastics).
Two more natural plastics of the trans-isoprene type will probably be exploited by Europeans in this time frame. Gutta-percha is produced by trees of the genera Palaqium and Payena, found in the Malay Archipelago. This was used as insulation for underwater electrical equipment and cables. (EB15/gutta-percha). Chicle (75% trans-, 25% cis) is derived from Manilkaea zapota, the Sapodilla Tree of Mexico and Central America. It’s now mostly used in chewing gum. (EB15/chicle; EA), but there are sources which indicate that it once was likewise used as electrical insulation (Wilcox 208).
Melamine-formaldehyde (Formica®) (OTL 1938) is harder and more water-resistant than urea-formaldehyde, therefore used in table and counter tops, and dinnerware.
Polystyrene is a hard, rigid thermoplastic used in packaging and electrical and thermal insulation. Its Achilles’ heel is its brittleness. The first impact polystyrene was made in 1927 by blending polystyrene with natural rubber. (Scheirs 18). Modern impact polystyrene is a blend of polystyrene with 5–10% polybutadiene rubber.
Styrene is compatible with many different copolymers. One that’s mentioned in EB15/Ion-Exchange Reaction is divinylbenzene; together they form a polymer which, after chemical treatment to introduce sulfonic acid or quarternary ammonium groups, is a very useful ion-exchange resin.
Polyurethane. Foams, coatings, fibers (Spandex is a segmented polyurethane, see EB15 for particulars).
Polyethylene terephthalate (PET) is a polyester thermoplastic; strong, stiff, transparent, chemical resistant, gas-impermeable. It’s used to make fibers, recording tape, bottles, conveyor belts. Mylar® is a PET film, stretched horizontally and vertically. Dacron® is a closely related polymer.
At least as of late 1634, polyester isn’t being produced yet; Nicki Jo Prickett of Essen Chemical says, of her cotton-polyester blend sweat pants, “we probably won’t have something like them any time soon.” Mackey, “Game, Set and Match” (Grantville Gazette 23). Since cotton is available from India, the polyester is the problem.
Polycarbonate. polyester; engineering thermoplastic. Transparent, high impact strength, high dielectric strength, cold resistance, but limited chemical resistance.
Polyphenylene sulfide. Engineering thermoplastic; useful up to 260oC (higher with glass filler), chemically resistant. With dopants, becomes conductive.
Polyvinyl chloride. A strong, tough thermoplastic. Pure PVC is used in conduits and siding; plasticized PVC (a copolymer with vinyl acetate) in flexible tubes and sheets, and wire/cable insulation.
Polyvinyl acetate (PVAc) is used in adhesives and paints, but it can also be used to make tough, flexible moldings. PVAc can be hydrolyzed to form polyvinyl alcohol, or reacted with formaldehyde or butyraldehyde to make polyvinyl formal or butyral.
Polyvinylidene chloride has low water/gas permeability, ideal for food packaging.
Polyethylene oxide (glycol). Depending on the degree of polymerization, this is a liquid or a low-melting solid (Carbowax®). It’s used as a thickening agent in cosmetics and pharmaceuticals. (Field 137).
Polypropylene oxide (glycol) has similar uses.
Butyl rubber. Excellent insulator; cold resistant; gas impermeable. Inner tubes. It cannot safely be blended with natural rubber. (Miles 299).
Polyphenylene oxide. Rigid engineering thermoplastic. Can be extended by blending with polystyrene.
Epoxy. Thermoset used as adhesive, coating, and, with fiber reinforcement, in structural parts.
Phenoxy. Same monomers as epoxy, but a thermoplastic. Transparent, ductile, heat and chemical resistant. Containers, pipes, adhesives, coatings.
Thiokol rubber. Excellent sunlight, oil and solvent resistance. Very low gas permeability. Lacquer and paint hoses.
Neoprene. The first commercially successful (1932) synthetic elastomer. The chlorine atom makes the monomer extremely reactive with free radicals, and indeed it polymerizes spontaneously. It is also compatible with many comonomers. (Kirk-Othmer 8:1045ff). Very good strength, abrasion resistance, flexion resistance; good oil/solvent resistance. Of the six major rubbers, the only one with good flame resistance. Used for oil-resistant footwear and balloons. Acceptable for transmission belting.
Unsaturated polyesters. The premiere matrix material for fiberglass composites. Their availability at this early stage is conditional on using fumaric acid isolated from natural sources.
The more common maleic anhydride-based polyesters will not be available until we have vanadium pentoxide catalysts, and hence I have slotted them into 1638–42.
Nylon 66. Polyamide. We think of nylon as a fiber (the hit song of 1943 was “When the Nylons Bloom Again,” a reference to wartime diversion of production from stockings to parachutes) but it’s also an engineering thermoplastic—strong, tough, abrasion resistant, flexible, and low friction. Nylon 6 has similar properties.
Nylon was admired by Heinrich Schütz in a letter to his mother in April, 1634. Carrico, “Motifs” (Grantville Gazette 21). It can be used, not just for scarves for Heinrich’s relations, but for making sails stronger, lighter, and less leaky than canvas. Cooper, “The Wind is Free: Sailing Ship Design, Part 1: Propulsion” (Grantville Gazette 21). Simpson has a nylon prosthesis (1633, Chapter 41), and the suspension system in up-time hard hats is made of nylon. Cooper, “Safety First: Industrial Safety in 1632, Part Two, Technical Aspects” (Grantville Gazette 18).
Aramid (ARomatic polyAMIDe.; Nomex®, Kevlar®). A high-performance thermoplastic used for ballistic fibers, composite reinforcement, machinery belts. Kevlar® aramid is “five times stronger per weight than steel.”
Polyimide. Engineering thermoplastic. Abrasion and chemical resistant, low friction, useable up to 250oC. Can be made in linear or network polymer forms.
Polymethyl methacrylate (Plexiglas, Lucite) is a transparent, rigid, hard, weather resistant thermoplastic. Think of it as a glass substitute, suitable even for aircraft windshields.
Polymethyl cyanoacrylate (Super Glue) is used as a contact adhesive (sold as the monomer and polymerized during use); e.g., to close incisions.
Vinyl ester resins. As a matrix material for composites, these are intermediate in price and performance between unsaturated polyester and epoxy resins. They also have very high chemical/environmental resistance. (azom.com).
Polyacrylonitrile. Wool-like fibers. Intermediates for production of carbon fibers.
Styrene-Acrylonitrile. Rigid, transparent thermoplastic, more heat and solvent resistant than polystyrene, intermediate in production of ABS.
Polytetrafluoroethylene. Thermoplastic. Resistant to chemicals, weather, heat, and cold; low friction. Used as high temperature electrical insulation, cooking surface, chemical equipment liner, bearings.
Polyvinyl fluoride. Thermoplastic. Transparent film.
Silicones. Thermoset. Can be prepared as a fluid (by stopping the polymerization), plastic (by use of a crosslinking co-monomer), or rubber (Miles 102). Heat and chemical resistant, water repellent, popular in high temperature electrical insulation.
AIS. A semisynthetic analog of ABS, see below.
Styrene-butadiene rubber. Excellent age and electrical resistance. Used for wire/cable insulation, tire sidewalls, and to some degree for tire treads. Less heat resistant than natural rubber, making it less useful for truck tires.
ABS. Hard, tough, heat-resistant thermoplastic, sometimes considered an engineering plastic.
Nitrile Rubber. Gas-impermeable, oxidation-, oil-, solvent- and heat-resistant. Gas and oil hoses.
Styrene-maleic anhydride Copolymer. EB15 says “used in automobile parts, small appliances, and food-service trays.”
Methyl rubber. The rubber substitute used by the Germans in WW I, and abandoned at the earliest opportunity. Its strength can be increased six-fold to 2400 psi (still inferior to natural rubber), by addition of carbon black, as was demonstrated by Whitby in the 1930s, but Grantville literature doesn’t mention this and indeed is universally critical of methyl rubber. No specifics are given, but besides the strength problem, it has very poor oxygen resistance, and the elastic form (W-rubber) became brittle in the German winter. (PSLC).
Evolution of Material Choices, by Application
For chemical reactor linings and piping, we are concerned with resistance to temperature extremes and a wide range of chemicals; strength is a secondary consideration. In general, the plastics containing fluorine or (to a lesser degree) chlorine exhibit the best chemical resistance, and the first of these we will develop will probably be polyvinyl chloride. Later, we will be blessed with polytetrafluoroethylene and polyvinyl fluoride. The best of the rest are probably the furan resins (vulnerable to oxidizing agents), polyphenylene sulfide, epoxy, nitrile, polyacetal, neoprene.
For mechanical parts, we value both tensile strength and impact resistance; we need to be able to fabricate to close tolerance and the polymer has to have the dimensional stability to maintain that tolerance. We will initially make do with polyacetals, but our options will expand to include polycarbonates and polyphenylene sulfides, then nylons. ABS, unfortunately, will be delayed by the problems of making polybutadiene, but a semisynthetic analogue could be made using natural rubber. If low friction is important, then polytetrafluoroethylene (Teflon) is also worthy of scrutiny.
For thermal insulation and mechanical shock absorption, we should fairly soon have polyurethane foam.
For optical parts, transparency is critical, but impact resistance is also important. We will have cellulose acetate first, then polycarbonate and polystyrene, and later polymethyl methacrylates.
Polymers have an important role to play as electrical insulation; high resistivity and dielectric strength are the key parameters. However, insulating use must be further subdivided into rigid versus flexible, and low frequency versus high frequency.
When mechanical strength is also important, we would probably make good use of phenolic resins, and later polystyrene, polyphenylene sulfide and epoxy. For simple encapsulation of discrete electronics, waxes, tars, gutta-percha, balata or chicle may also be of value.
For wire and cable, the best known flexible plastics are PTFE, polybutylene, LDPE, some polyurethanes, and plasticized PVC. Rubbers, by their nature, are also flexible. We will probably be using natural rubber and plasticized PVC in the mid-1630s for low-frequency cable.
To be suitable for high frequency use, we want a low “dissipation factor” at frequency, and preferably also a low dielectric constant. For rigid applications, we can use phenol-formaldehyde (dissipation 0.04@1MHz), polyacetal (0.008), polystyrene (0.0003), or even, if mechanical strength is unimportant, wax (0.001). I fear that for high frequency cable, initially we will be making do with tars (loss tangent 0.04) and natural rubber (0.03), and then with plasticized PVC (0.006). PTFE (0.0003) or silicone (0.002) will hopefully become available by the late 1630s.
It’s perhaps worth noting that cables may have both an inner layer as electrical insulation and an outer layer to protect the inner one from the elements.
Industrial Polymerization Methods
Older textbooks classify polymerization reactions as being addition or condensation polymerizations. Addition polymerization is usually exothermic, and the reactors must be designed to dissipate the heat. Otherwise, the retained heat will increase the speed of the reaction, which will produce more heat, which . . . well, the result wouldn’t be pretty. Condensation polymerization is usually endothermic, so the reactors must efficiently supply heat to the reaction so it doesn’t peter out. (EB15).
EB15 describes five basic industrial polymerization methods, which are classified according to how the monomers are brought together.
Bulk (Melt). Without any solvent or dispersant. Simple, but heat control and stirring can be difficult. Also, monomer purity is quite important. (Bahadur 57).
Solution. Alleviates the problems of bulk polymerization, but the solvent can react with the polymer and also be difficult to remove from the polymerization product. Hence, it is preferred for products sold as liquids, such as some adhesives and coatings. Note that gas phase monomers can be dissolved in a solvent.
Suspension. If the monomer won’t dissolve, it can still be suspended in a liquid by stirring. The polymer is produced in a granular form.
What Grantville literature doesn’t mention is what happens if you don’t stir sufficiently. If there isn’t enough agitation, and the polymers haven’t polymerized enough so that their glass transition temperature is above the working temperature, the particles coalesce, and you get a “runaway” bulk polymerization. For similar reasons, suspension polymerizations cannot be run continuously. (Kirk-Othmer 19:899).
Emulsion. A detergent is added to emulsify the monomers. This works well with free radical-initiated reactions, if you want a high molecular weight polymer, because the polymerization reaction isn’t terminated as easily. Finding the right detergent can be tricky and purification is difficult, so this is best for producing coatings and adhesives that can be used as emulsions.
Gas-Phase. If the monomers are gaseous, they can be reacted under pressure. The polymer, because of its molecular weight, will be a solid.
CCD715 gives some exemplary conditions: Bulk, 150oC; Solution, -70 to 70oC; Suspension, 60–80oC; Emulsion, -20 to 60oC; and Gas-Phase, high pressure and temperature (200oC).
Another technique, not mentioned by EB15, is interfacial polymerization. One monomer is dissolved in solvent A, the other in solvent B. If the two solvents can’t mix, and the monomers can’t self-polymerize, then the reaction can occur only at the boundary of the two liquids.
Fabrication of Polymeric Materials
A polymeric material is either a pure polymer, or a material that consists predominantly of polymers but to which other substances have been added. Polymeric materials can exist in a variety of forms:
2-dimensional: sheets, fabrics and coatings
3-dimensional: solid parts; foams; liquids
Polymeric materials can comprise plastics or elastomers. Plastics can be thermoplastics or thermosets. Thermoplastics (linear and branched polymers) soften and melt when heated, and will dissolve in certain solvents. They can be reheated and deformed again and again.
Thermosets (networked and crosslinked polymers) suffer chemical reactions when heated so that once they cool they are permanently set (“cured”) in a particular shape. They also are then insoluble. If reheated, they decompose. They are not suitable for use as fibers. (Wittcoff I:163). They tend to be harder and more brittle than thermoplastics, and they are more difficult to process into a desired shape. (Wittcoff II:40).
The commercially important thermosets are phenol-, urea- and melamine-formaldehyde, epoxy, unsaturated polyester (used to make fibreglass), furan resin, and some polyurethanes. Most elastomers also behave like thermosets once they have been vulcanized.
EB15 and EA outline the basic methods of forming a polymer into the desired shape. The sine qua non is causing the polymer to flow. You have four choices: melt it, dissolve it, suspend particles of it in a liquid, or put it under sufficient pressure that it flows even at room temperature.
Extrusion involves putting the polymer, heated or not, under pressure and forcing it out through a die. Extrusion can be driven by a piston (ram), but screws are more popular. Screw extruders were built in the 1840s to coat copper conductors with gutta-percha.
If you want to make fibers, you use a die with lots of small holes (spinnerets). If the die has a narrow rectangular slot, you extrude a sheet, and if it has an open ring, you get a pipe. The key point about extrusion is that the object will have a constant cross-section.
While the technology is a fairly straightforward one, it has its subtleties, “the specific design of the screw(s) having an enormous effect on the degree of mixing and the temperatures that are generated inside the chamber.” (Teegarden 160). About 60% of all plastics are extruded.
In injection molding, the extruder, instead of forcing the material out through a die, forces it into a mold. The mold must be clamped to resist the extrusion pressure. The first injection molding machine, patented by the Hyatt brothers in 1872, used a ram extruder, whereas screw injection was introduced by Hendry in 1946. (Wikipedia/Injection Molding).
Injection molding took advantage of the reformability of thermoplastics. Historically, injection molding was an adaptation of die casting of metals. The first injection molding machines were hand operated, but the first fully automatic machine appeared in 1935. An aspirin box could be produced in 15 seconds by injection molding, as opposed to 15 minutes by compression molding. The machines of 1930 could only inject a half ounce “charge,” but this was scaled up to 16 ounces by 1939. (Meikle 80ff).
In blow molding, air is blown into a plastic hollow tube (parison), which thereby expands against the inner surface of a mold. I have seen it asserted that this technique was known to the ancient Egyptians. Of course, the down-timers don’t have any blow-molding machines, but we can certainly compress air and feed it into a parison, mechanically. Blow molding machines were used in the early 1900s to make celluloid bottles, dolls and ping-pong balls.
Thermoforming takes a sheet of polymer, heats it until it softens, and rams it into a mold. Primitive forms of this method (heating tortoise shells in hot oil, or tree bark in hot water, and then forcing then into a mold to make a container) date back to ancient times. (Throne, Understanding Thermoforming 2). Thermoforming machines were developed in the 1870s.
In compression molding, the polymer (usually preheated) is first placed into the mold, and then pressure is applied (as opposed to pressure forcing it into a mold as in injection molding or thermoforming).
While not discussed by EB15, the up-timers are probably aware of the possibility of hand lay-up or spray-up with certain plastics (notably unsaturated polyesters and epoxies) as these are part of the home hobbyist arsenal. Polyesters could be laid-up by hand because they didn’t give off water or gas during formation. (Meikle 158).
Thermoset plastics (the phenol-, urea- and melamine- formaldehyde copolymers, epoxies, and unsaturated polyesters) are usually compression molded, whereas injection molding is the most common process for thermoplastics.
Foams. The holes in a foam are made by gas bubbles. If you inject an inert gas (air, nitrogen, carbon dioxide, etc.) under pressure directly into the polymer melt, the foam will have an open cell structure. Or you can use a volatile liquid that evaporates to make the gas. If instead a foaming agent, which decomposes at the melt temperature to produce a gas, is incorporated into the resin, you get a closed cell structure. The foam may be flexible or rigid, depending on the plastic chosen. The most commonly foamed plastics are vinyls, polystyrene (Styrofoam), polyethylene, phenolics, silicones, cellulose acetate, and polyurethanes (EB15/foamed plastic). Foam (sponge) rubber is made by whipping the latex into a froth containing about 85% air, or by using a foaming agent.
Fibers. If the polymer is being formed into a fiber, it will be melted and spun, or dissolved, gelled or emulsified, and spun. (EB15/MMF) The spinning is really an extrusion of the polymer through small holes. During or after spinning, the fibers are drawn (stretched) to orient the polymer chains and thereby increase their tensile strength. Drawing can be cold or hot.
There was a long gap between when polyacrylonitrile was first made (1941) and when it became commercially significant (mid-fifties). The problem was that it can’t be spun from the melt, because it isn’t heat-stable enough. It took time to find a suitable solvent (dimethylformamide). EB15 mentions that such a “spinning solvent” exists but doesn’t identify it. Hence, unless we are luckier than the DuPont chemists of the Forties, we will probably make copolymers of acrylonitrile and other monomers, rather than pure polyacrylonitrile.
The drawn fibers may also be textured to change their feel, and of course they are often dyed. Dyes may be introduced before, during or after spinning.
Processing time is of critical importance in determining whether a particular polymer/fabrication method is successful in the modern world (the hardening time for polylactic acid had to be reduced from a minute to few seconds in order to make PLA production economically viable) but I assume it will not loom so large in the first decade or two after RoF.
While the polymer is the main ingredient of a plastic or rubber, other ingredients are usually added to it to improve its properties.
Fillers. Fillers for plastics come in two principal categories, reinforcing agents to increase strength (finely divided silica, carbon black, glass fibers) and extenders (like sawdust, to reduce cost). Mica and asbestos are added to increase heat resistance. (EA). Fillers may be particulate (spheres, cubes or flakes) or fibrous. A plastic with fibrous reinforcement is usually called a “composite,” see below.
Stabilizers. To protect against oxygen, you may need antioxidants such as “hindered phenols” (butylated hydroxytoluene) or tertiary amines. Some plastics will require protection against heat, sunlight, or microorganisms.
Pigments and Dyes. These include titanium and zinc oxide (white) carbon (black), and oxides of iron or chromium.
Plasticizers. These are low molecular weight organic compounds that soften the polymer. The number of commercial plasticizers rose from 56 in 1934 to 414 in 1987. (Deanin 167). Variety is good, because a plasticizer suitable for one polymer might not work with another.
One standard plasticizer is dioctyl phthalate. Plasticizers are frequently added to PVC.
Flame retardants. These are typically rich in chlorine or bromine.
Foaming agents. These decompose at the molding temperature, evolving large quantities of gas. They include isopentane and azidodicarbonamide.
If additives are used, they must be mixed into the polymer in some way. If both are liquid, this is trivial. Otherwise a powerful blender is needed.
It is very important to recognize that additives can drastically alter the characteristics of a polymer.
Polymers may be blended together, just as metals are alloyed together. If they are miscible—that is, each dissolves in the other—then the properties of the blend are usually weighted averages of those of the component polymers.
If the polymers are immiscible, then unless precautions are taken, they will separate under stress. This can be avoided if the two polymers are “embedded” in a network formed by a third polymer. Alternatively, a homopolymer of monomer A and immiscible homopolymer of monomer B are blended with a block or graft copolymer of monomers A and B. (EB15/IPC).
Polystyrene is usually blended with 5–10% polybutadiene to make high-impact PS.
A composite is a “macroscopic” combination of two or more constituent materials to form a new material, in which the original constituents remain distinguishable. This can result in very different properties. Unfilled Bakelite has a tensile strength of 5000 psi, but wood flour filler can increase it to 7000 (Miles 338). A typical epoxy-boron fiber composite (70% fiber) has a flexural strength of 300,000 psi. (EA/Composite Materials). In contrast, a simple epoxy resin has a flexural stress at fracture of at most 23,000 psi. (MCA 38). (Just to complicate matters further, a constituent material of a composite may itself be a composite.)
The composite may comprise matrix material which surrounds a filler material; the latter may be particles, flat flakes or fibers. Or a skeleton or honeycomb of one material may have its open spaces filled by a second material. Or you may have a laminate, in which alternating layers are of one material or another, as in the case of plywood.
The particles of particulate composites include glass spheres, calcite, kaolin, mica, talc, carbon black, wollastonite, feldspar, metal or metal (zinc, iron, titanium, antimony) oxide powders, and wood flour. Particulate composites are almost as old as synthetic plastics themselves. Note that many synthetic rubbers require carbon black for strength.
Fibrous composites are of particular interest. The fiber provides strength, and the matrix protects the fibers from environmental attack. Pure Bakelite is brittle, but very early in its development, “wood flour” (short cellulose fibers) were added to toughen it. (Kirk-Othmer 7:2).
The fiber content of glass fiber-reinforced plastic (GFRP) is such that it might more aptly be described as plastic-protected glass fibers. (Note we usually make structures out of GFRP, not “fiberglass”; “fiberglass” strictly speaking refers just to the bundle or mat of glass fibers.)
When the reinforcement is a fiber or yarn (twisted fibers), the orientation matters; they may be unidirectional, bidirectional (as in a fabric), two-dimensionally random or completely random. Composites with controlled orientation will probably be stronger than a random orientation composite in some directions and weaker in others.
Another important consideration is the length of the fiber, long fibers being better (sometimes!) at resisting loads but making it harder to fabricate the composite into a desired shape. It naturally is possible to use a combination of fibers of different materials or lengths.
The strength and stiffness of the composite are intermediate between those of the matrix and the fiber. It’s difficult to achieve a fiber content of even 50%. (Gordon 187) but this depends on fiber orientation and length, and the fabrication method. With hand lay-up on glass fibers, 20–40% may be expected, but some methods yield up to 80% (Kirk-Othmer 7:31).
Here, we are concerned with composites in which either the matrix or the fibers are plastics. The most commonly used plastic matrices are epoxy, polyester, and vinyl ester resins. Phenolics were once popular, but are now used mostly when their flame and heat resistance, and low cost, outweighs their inferiority in strength. (Kirk-Othmer 7:29).
The fibers will usually be made of a material with a high tensile strength. Glass and carbon fibers were discussed in Cooper, “Better Foundations, Part 1: An Introduction to Concrete” (Grantville Gazette 19). However, it’s important not to get fixated on them. Other possibilities include vegetable (wood, bamboo, sisal, banana, hemp, cotton, flax, jute, coir, ramie, etc.), metal, asbestos, plastic (polyacrylonitrile, nylon, polyethylene, polypropylene, aramid), boron, ceramics, and basalt.
In 1943, Henry Ford showed off a plastic car with fourteen composite panels that were 70% a combination of southern-pine, straw, hemp and ramie fibers, and 30% Bakelite. The panels were quarter-inch thick, but allegedly had an impact strength ten times that of steel. (Stidger).
A thorough discussion of when the various fibers can be made in the new time line, and what their advantages and disadvantages are, is beyond the scope of this article, but I hope to address them in a mini-article in the near future.
The fiber and matrix must be compatible; that is, the matrix must adhere to the fiber so that stresses on the matrix are transferred to the fiber, which is better able to withstand them. The fiber may receive a mechanical surface treatment or be coated with a chemical coupling agent to improve adhesion.
There are several methods of fabricating a FRP composite. Resin may be hand brushed or sprayed on a fiber mat. In transfer molding, the mold is filled with the fibers and then the resin is injected into the mold. In compression molding, the pre-impregnated composite is laid in the mold and then squeezed. In filament winding, fibers or tape (optionally precoated with resin to make a “prepreg”) are drawn through a resin bath and then wound onto a mandrel. In pultrusion, a fiber bundle is pulled through the resin matrix bath and then through a heated die to make a composite with a constant cross-section.
In June 1633, a composite of glass fiber cloth and viscose was tested by Markgraf Aviation as a possible fuselage material. A GFRP wasn’t used on their first plane, the Mercury, because the fiberglass was too expensive. A technical note implies that fiberglass was used in their second plane, the Jupiter (nicknamed, “The Monster”), but the story doesn’t actually say so. In January, 1634, when it was being built, there were arguments about which resin to use in making the composite, so viscose clearly wasn’t the only one under consideration. And we don’t know what the final choice was. Huff and Goodlett, “The Monster” (Grantville Gazette 12).
In view of the polymer timeline, the most likely resins are agglutinated casein, casein-formaldehyde (galalith), viscose (rayon, regenerated cellulose), cellulose acetate, and phenol-formaldehyde (Bakelite).
Agglutinated casein softens and ultimately dissolves if exposed to alkali. With galalith and viscose, water absorption is severe (for galalith, 33% per Scherer 115 and 7–14% per Brady 152; 100% for rayon, Lewin 788), and impairs dimensional stability. In contrast, it’s 2–5% for cellulose acetate and under 1% for Bakelite (MCA). Also, I worry about the casein and cellulose-based plastics being biodegradable—they will be exposed to moist air and fungi, and I know that the casein glues once used in aircraft deteriorated (Gordon 158), as did the cellulose fibers in a paper-based phenolic used in a German WW I aircraft (182).
While these problems could be alleviated by coating the GFRP with a waterproof paint, possibly containing a biocide, that would certainly add to the production cost. Glass fiber-Bakelite composites are known to be functional (183), and Bakelite was made by Dr. Gribbleflotz’ laboratory in 1634. And perhaps Markgraf reproduced it independently earlier. So I favor Bakelite as the first matrix material.
Recycling Up-Time Polymers
Many polymeric materials contain dyes, fillers and other additives, and it isn’t easy to remove them. If waste thermoplastics are in pure enough form, they can be reground and reused.
This is most likely to be possible with beverage bottles and plastic bags. And it makes sense to remove labels, if possible.
Such recycling presupposes that we can identify the plastic. This is simplest if the waste bears the standard resin identification codes: 1 PET, 2 HDPE, 3 PVC, 4 LDPE, 5 PP, 6 PS and 7 Other. If not, then we can try measuring its density, and there are also some simple chemical tests. Thermosets (which includes most rubbers) can’t be recycled.
My suspicion is that the post-RoF value of up-time plastic articles qua up-time “relics,” if nothing else, will be much greater than the value of the raw plastic that might be obtained from them by recycling.
World annual per capita plastics consumption rose from 0.03 kilograms in 1920, to 0.16 in 1940, 2.2 in 1960, 13.3 in 1980 and 26.3 in 2000. In 2000, it was 102 kg in the United States and 133 in Germany, but only 1 kg in India. (Elias 8).
The price of modern plastics is extremely dependent on that of oil. The plastics of the new timeline are derived from coal and natural feedstocks, and hence their prices will be “pegged” accordingly.
We want to minimize the number of synthetic steps in monomer production both to maximize monomer purity and minimize costs. A useful rule of thumb is to figure that each synthesis step produces an intermediate whose price is 2–3 times the raw materials cost. (Wittcoff).
In the course of the first efforts to produce synthetic rubber, the initial obstacle was the high cost of monomer production; Hoffman’s 1912 process took six steps to go from p-cresol to isoprene (Morawetz 64). On the other hand, the “dumping” of war supplies of phenol after WW I contributed strongly to the early commercial acceptance of Bakelite.
Bear in mind that prices of products produced in R&D quantities are usually very high. As plastics become more popular, their prices should decrease as a result of economies of scale. The price of the newly introduced “furfural plastic” dropped from $65/pound (1917) to $3 (1921)(Weekly Commercial News). Polyethylene was introduced at $0.40 but fell quickly to under $0.10—the ethylene monomer was available for less than $0.02 from petrochemical production.
Another useful guideline is that the price per unit is proportional to the -0.4th power of the annual production (Elias 243). Hence, if volume doubles, the price will drop by 24%.
In 1984, polypropylene and PVC had the same cost per weight as polyethylene; ABS 2.5x; polyester and furan, 3x; nylon 66, polyacetal, and polycarbonate, 3.5x; epoxy 4x, polyphenylene sulfide 10x, and PTFE, 12x. (Cornish, 98). Relative prices are somewhat different if costs are figured on a per volume basis, with PTFE 26 times polyethylene. However, polyethylene and polypropylene will not be relatively cheap within the first decade of the new time line.
The pricing of plastics before petrochemical feedstocks became dominant may be of interest.
Table 5-2 Selected Pre-WW II Polymer Prices
Inflated to 2000
celluloid (volume 213,000 pounds)(Meikle 17)
celluloid (volume 8,400,000 pounds)
galalith (Industrial Engineering)
fancy buttons, galalith (BFDC)
fancy buttons, celluloid
Bakelite (Meikle 46)
Bakelite (Meikle 56)
cellulose acetate (volume 2,800,000 pounds)(Meikle 82)
cellulose acetate (volume 20,300,000 pounds)
polystyrene (VICTRON)(Meikle 89)
natural rubber, high 1900–1940 (Morawetz 65ff)
natural rubber, low
nylon (volume 2,600,000 pounds)(Hounshell 1972)
(1) 1913+ conversion per BLS CPI-U, pre-1913 per Sahr.CPI.
Natural rubber prices were mostly in the $0.10–1.00/pound range during the period between the world wars. In peacetime, when it was at the low end, the interest in synthetic rubber was purely academic (PSLC). When neoprene was introduced, it sold for $1.05/pound, but natural rubber was then 3–5 cents/pound (Morton 339), relegating neoprene to niche (solvent resistance) markets.
PART B: CONSTRUCTION OF THE TIMELINE; SELECTED PRODUCTION PROBLEMS
Chemically speaking, polymers are large molecules formed by the combination of many units of one or a small number of chemical building blocks, called monomers. The polymer thereby consists of numerous repeating units derived from the original monomers.
The introduction of a synthetic polymer may be delayed because:
1) Grantville literature does not reveal how to make one of the required monomers, or the known production methods require starting materials, reagents, catalysts, or process conditions that we can’t yet provide, or can provide only at a cost that renders the process uneconomical;
2) there are problems in controlling the polymerization so we achieve a reasonable polymerization time, or so we limit the degree of polymerization so we obtain a processible polymer; or
3) the properties of the polymer are dependent on controlling its stereochemical structure, and to achieve it, we need a catalyst or initiator that we don’t know how to make or which we cannot make cheaply enough.
Synthetic Polymer Production
Plainly, the first step toward duplicating in the new time line a polymer of the old one is figuring out which monomers it is based on. Often, that’s pretty trivial; polymers are named on the basis of the monomers that were polymerized to make them. Thus, “polypropylene” is the product of the polymerization of propylene. However, if you just know a polymer by its tradename or trademark, such as nylon, Orlon®, or Kevlar®, then you will need to do some library research.
For example, vinyl ester resins are alluded to in EB15/airplane, but the monomers aren’t identified. (One class of vinyl ester resins are essentially acrylate or methacrylate-modified epoxy resins, and hence should be producible once we have solved the problems of making epoxy and polymethylmethacrylate resins.)
Next, to make the polymer we need to (1) be able to make the required monomer(s) and (2) work out the polymerization chemistry. Table 5-1 shows which chemicals are used as monomers in the manufacture of which polymers. Some can be used to make more than one polymer, and of course any heteropolymer requires at least two different monomers. Some polymers are made, not by polymerizing one or more monomers, but by modifying an existing polymer.
Being able to prepare the component monomers is just part of the battle. The process conditions (pressure, temperature, catalytic agents, etc.) determine whether polymerization occurs, the degree of polymerization achieved, the regularity of the arrangement of the units of a heteropolymer, the stereospecificity of the repeating units, and so on.
The degree of polymerization strongly affects the polymer properties. To achieve a high degree of polymerization you need very pure reagents (monofunctional reactive impurities would terminate the chains) and precise reaction proportions of the monomers, and you must drive the polymerization to completion. To achieve stereoregularity or stereospecificity you usually need special catalysts or initiators.
Monomer Availability Problems
In the Thirties, the American organic chemistry industry went through a revolutionary change, switching from coal tar to petroleum as its principal feedstock. By the Fifties, this revolution had spread even to coal-rich, petroleum-poor Germany.
As a result of this changeover, modern plastics production is dependent on petrochemical feedstocks, and, when modern Grantville literature discusses how to make organic chemicals on an industrial scale, the emphasis is on petrochemical processes.
Even in the late twentieth century, most (90%) of the petroleum produced was consumed as a fuel. It was only because the production was on such a huge scale that the organic chemical industry could rely on petrochemicals. And it did so, of course, because of cost.
Price is an important factor in whether plastics are used, and the monomers used to make the major plastics were, in 2000, all petrochemicals. Because of the limited supply of crude oil in the 1632verse, and the large captive demand (unused cars and trucks) for gasoline, the price of petrochemicals will be high. In 163x, we are going to have to build the plastics industry, like the more general organic chemical industry, on the isolation and transformation of chemicals from coal tar, biological sources like animal fats and ferments, and inorganic sources such as carbides.
The principal heteropolymers, and their component co-monomers, are listed in table 5-3. The “C#” after the monomer is the size of the largest carbon moiety in the monomer, and is provided because the next table arranges the monomers by C#.
Table 5-3 Heteropolymers
formaldehyde (C1), phenol (aro)
formaldehyde, urea (C1)
formaldehyde, melamine (misc)
furfural (C4), phenol
phosgene (C1), bisphenol A (aro)
ethylene dichloride (C2), sodium sulfide (C0)
ethylene (C2), propylene (C3) and a diene (C4+)
ethylene glycol (C2), terephthalic acid (aro)
ethylene glycol (C2), methyl terephthalate (aro)
epichlorohydrin (C3), bisphenol A (or glycerol or cashew nut phenol)
acrylonitrile (C3), butadiene (C4)
acrylonitrile, styrene (aro)
acrylonitrile, butadiene (nominally) and styrene
glycerol (C3), phthalic anhydride (aro)
butadiene (C4), styrene (aro)
isobutylene (C4), isoprene (C5)
hexamethylenediamine (C6), adipic acid (aro)
dichlorobenzene (aro), sodium sulfide
aldehyde, aromatic alcohol (phenol, resorcinol, cashew nut phenol)
polyalcohol, polyfunctional acid or acid anhydride
saturated diol (e.g., diethylene glycol (C2) or propylene glycol (C3)) and unsaturated acid or acid anhydride ( fumaric acid (C4) or maleic anhydride (C4))
aramids (includes Kevlar®)
diamine, terephthalic acid or derivative
Table 5-4 outlines my proposed non-petrochemical routes for obtaining various monomers. The table lists the monomer, the “ultimate” building block, the feedstock from which the building block is obtained, the number of steps (S#) needed to obtain the monomer from the building block or (+n) another monomer in the table. When it isn’t obvious which polymer the monomer is used to make, this is indicated in parentheses.
Table 5-4 Proposed Non-Petrochemical Routes of Monomer Availability
(monomer boldfaced if forms homopolymer, italicized if forms heteropolymer with another monomer)
wood; fermentation discard
Polyacetal, see below.
carbon monoxide and activated charcoal
a) uric acid
b) carbon dioxide
a) urine, guano
b) limestone; fermentation
a) ethyl alcohol
a) fermented starch
b) limestone, coke
from acetylene, ethylene, or ethylene dichloride
b) acetic acid
a) limestone, coke
a) limestone, coke
b) ethyl alcohol
a) natural gas
need hydrogen fluoride and antimony chloride catalyst
a) from ethylene oxide
b) from formaldehyde and carbon dioxide
coproduced with ethylene glycol from ethylene oxide
fusel oil (fermentation discard)
b1) acetone and b2) HCN
a) fusel oil
b2) coke oven gas, some plants
a) needs fancy catalyst
b) cuprous chloride catalyst
no synthesis information in Grantville literature, but made historically from acetone by way of pinacone. (Martin 366; Morawetz 65).
a) initially from acetylene, b) later from butadiene
from acetone, HCN and methanol
from acetic acid, potassium cyanide, methanol, formaldehyde
byproduct of production of phthalic anhydride, see below
fats, fusel oil
natural rubber, turpentine, various oils
Nylon 6,6: need adipic acid
a) from adipic acid (comonomer, see below) in 3 steps
b) from 1,3 butadiene in 3 steps
c) from furfural (in cellulosic waste) in 5 steps
d) from formaldehyde, acetylene and cyanide in 5 steps
also a precursor to a diisocyanate used to make polyurethane.
b) petroselinic acid
a) coal tar
b) parsley seed oil
also possible precursor to hexamethylenediamine
from acetic acid and phenol
from ethylene and benzene
dichlorobenzene (polyphenylene sulfide)
from acetone and phenol
b) calcium carbide (see below)
various substituted silanes (silicone)
silica (sand, glass waste), coke
from toluene (coal tar) and phosgene
one of many polyisocyanates used with polyol to make polyurethanes
C1 Chemistry. The C1 monomers should be fairly easy to make, but to use them to make a polymer, we need additional chemicals.
C2 Chemistry. The use of acetylene as a synthetic intermediate was discussed in Cooper, “Industrial Alchemy, part 3.” Acetylene is made by fusing lime and coke in an electric arc furnace (resulting in calcium carbide) and then adding water.
A miner’s lamp that reacted calcium carbide and water to produce (and burn) acetylene first appeared in canon in January, 1635, so I have assumed that acetylene is available in 1634. Acetylene is the key C2 intermediate (as it was in WW I Germany) and makes possible the production of Thiokol rubber (1636–8), polyvinyl chloride (1635–7), polyvinylidene chloride, polyethylene oxide (1635–7), and one of the comonomers used to make polyethylene terephthalate (1635–7), polyurethane (1635–7), and unsaturated polyester (1636–38).
If for some reason we couldn’t use acetylene, the logical C2 building blocks would be ethyl alcohol or acetic acid, both available from winemaking and wood processing.
While polyethylene and polytetrafluoroethylene are derived from C2 monomers, they are late in my chronology, PE because of polymerization problems, and PTFE because of the general reluctance of sane chemists to work with hydrogen fluoride.
C3 Chemistry. C3 monomers aren’t too difficult to make, but there are polymerization problems facing polypropylene and related heteropolymers.
C4/C5 Chemistry. Chloroprene and the acrylics can be built by combining C1-C3 building blocks, putting the corresponding polymers in the late 1630s time frame.
The polymers based on butadiene (C4) and isoprene (C5) face something of a double whammy, as the older methods of making the monomers (from non-petrochemical sources and without fancy catalysts) aren’t well documented in Grantville literature, and there are also problems in achieving the stereochemical control during polymerization necessary to use them to make a general purpose synthetic rubber. These problems impact our ability to make polybutadiene, polyisobutylene (which comprises some isoprene), polyisoprene, styrene-butadiene, butadiene-acrylonitrile, and ABS polymers.
Lebedev apparently obtained butadiene from ethyl alcohol circa 1910, and in WW II that route accounted for 40% of American butadiene production, and one with acetylene as the precursor for 60%, but details aren’t stated. (EB15/elastomer, butadiene). Reconstructing one of these methods could make a big difference to the polymer timeline for all polymers with butadiene or derivatives as a monomer.
Doing some sleuthing outside Grantville literature, it appears that the Lebedev process was a single stage, gas phase reaction at 400oC, 0.25 atm. pressure, over a zinc oxide-alumina catalyst. It had a yield of 50%. (Petroleum Chemicals Industry 231). Another ethanol-based process was developed by Ostromislensky, reacting it with acetaldehyde over an oxide catalyst at 360–440°C (Comyns 189). The acetylene-based method probably involved heating acetylene with ethylene (Martin 366c) or formaldehyde (DeBell 507).
Another option would be to isolate Clostridium acetobutyricum, or a similar organism, that converts starch into butyl alcohol. This is treated with HCl, producing various chlorides, and the correct one must be isolated by distillation and then treated with soda lime.
Otherwise, we will probably make butadiene by reacting acetylene and formaldehyde under high pressure to get 1,4-butanediol, and then dehydrate the latter with acid and heat. This will be a more cumbersome process to develop, and probably on the expensive side.
As for isoprene, while Tilden isolated it from turpentine oil in 1882 (PSLC), the typical yield was only 6% (Martin 366d). Perkin recommended converting amyl alcohol (from fusel oil) to isoprene by methods analogous to that for converting butyl alcohol to butadiene.
The C4’s maleic anhydride and fumaric acid are obtained by vanadium pentoxide-catalyzed oxidation of benzene. Unfortunately, I am not expecting vanadium to be available until the 1637–39 time frame. This in turn will probably delay the appearance of the unsaturated polyesters.
Maleic anhydride is also available as a byproduct (5%—Bjorksten 22) of phthalic anhydride production, and can be converted into fumaric acid. Unfortunately, phthalic anhydride requires the same catalyst.
However . . . EB15 does mention that fumaric acid can be used instead of maleic anhydride, and EB11/Fumaric and Maleic Acids states that it’s found in “fumitory (Fumaria officinalis), in various fungi (Agaricus piperatus, &c.), and in Iceland moss.” Fumitory is referred to in herbals from the sixteenth century and should be recognizable and available at least in Britain. (EB11/Fumitory). The fumaric acid content of fumitory is variously reported as 6.156% (Watts 742) or 0.156% (Gmelin 23) of the fresh herb. Iceland moss is found in many parts of northern Europe, and is probably known to down-timers at least in Iceland, but I haven’t been able to ascertain its fumaric acid content.
C6/Aromatic Chemistry. Coal tar feedstocks makes it relatively easy to make aromatic monomers, and by some further chemical shenanigans, the linear C6 hexamethylenediamine.
In the Appendix I go into detail as to what Grantville literature has to say about how to make each of these monomers, sometimes by several different routes.
In Table 5-5 I am proposing the following monomer availabilities:
Table 5-5: Optimistic Monomer Availability Timeline
C1: formaldehyde, phosgene
C1: urea. C2: ethylene. C3: glycerol, glycolic acid, lactic acid. Aro: phenol, xylenol
C2: ethylene oxide, ethylene dichloride, ethylene glycol, vinyl chloride, vinyl acetate. C3: propylene. C4: isoprene (from natural rubber); isobutylene (from fats), furfural. Aro: bisphenol A, styrene, terephthalic acid. Misc: melamine, other polyols
C2: vinylidene chloride. C3: epichlorohydrin, propylene glycol. C4: butadiene, chloroprene (from acetylene); fumaric acid (from fumitory?). C5: synthetic isoprene; C6: adipic acid, hexamethylenediamine. Misc: some diisocyanates
acrylonitrile, methacrylates and cyanoacrylates; caprolactone and caprolactam; hydroxybutyrate
tetrafluoroethylene; synthetic 1,3-butadiene; 1,4-butanediol; synthetic fumaric acid, maleic anhydride, phthalic anhydride
synthetic isobutylene (by petroleum cracking)
Experts say that “more time is wasted in polymer chemistry by impure monomers than for any other reason.” (Green 125).
M&B 261, discussing free-radical polymerization, warns, “since even traces of impurities, acting as chain-terminators or chain-transfer agents (interrupting one chain to start another), can interfere drastically with the polymerization process, the monomers used are among the purest organic chemicals produced.”
The problem is worse with a linear polymer, because each chain has only two “growing ends.” If a monofunctional molecule impurity (one with only one relevant reactive group) reacts with one end of the chain, that “kills” that end, reducing its activity 50%. If a second such impurity comes along and reacts with the other end, the chain is “dead,” wasting all of the monomers that have already been strung into that chain if it isn’t long enough to achieve the desired properties.
Bifunctional impurities don’t prevent chain growth, but they can create mechanical or chemical weak points in the polymer. Trifunctional impurities do that, too, and they also start branches, which aren’t desirable if you’re trying to make a linear polymer.
How pure is pure enough? Modern expectations are 99–99.9% purity, depending on the polymer (Arlie, 12, 14, 39, ArlieCT, 26, 77). Moreover, there are especially stringent limitations on specific impurities, for example, of 3 ppm on acetylene in making polyethylene, or 1 ppm on oxygen and sulfur in making polypropylene.
Hence we may find that even though we can produce the necessary monomer, we cannot provide it in the necessary purity.
Since nylon is a linear polymer, polymer growth could easily be halted by reaction with a monocarboxylic acid or monoamine impurity. The solution was to purify the monomers by crystallization of a salt of the acid and the diamine. (Teergarden 126).
Commercial production of polystyrene began in 1931, using styrene derived from ethylbenzene. (Wunsch 7). EB15 says that it wasn’t possible to prepare a commercially acceptable polystyrene (there were problems of brittleness and cracking) until 1937, when the monomer was obtained in greater purity.
Step growth polymerization involves reaction between monomers that react on their own. Heteropolymers are typically “grown” by step growth; the “X” group of a “di-X” reacts with a “Y” group of a “di-Y” to form a linear polymer. A step-growth polymerization is usually a condensation (the reaction making nylon 6,6) but may be an addition (the one making polyurethane). Catalysts and initiators are not needed, but the polymers grow slowly as the reaction progresses.
Chain growth polymerization requires an initiator. The initiators are not considered catalysts because they are consumed by the reaction. The initiator reacts with the monomer or nascent polymer to form a reactive intermediate. High polymer is formed quite rapidly; so increasing the duration of the reaction simply increases yield (Teergarden 89).
One important class of initiators are compounds that break down to generate a free radical. Peroxides, such as benzoyl peroxide, fall into this category. The free radical that attacks the monomer, converting it into a reacting species that attacks a second molecule of monomer. The resulting dimer is thus itself activated, attacking a third molecule of monomer. And so on. The reaction terminates if two activated molecules combine.
Cationic initiators, such as sulfuric acid, transfer a hydrogen ion to the monomer in order to activate it. This method has been used to polymerize isobutylene (Kirk-Othmer 19:901).
Anionic initiators, such as methyllithium, add to the monomer to form a reactive anion. Or, like sodium metal, they transfer an electron to the monomer to make an anion that is also a radical. Anionic initiation leads to an essentially monodisperse (polymer molecules of equal length) product; in the absence of impurities, the polymerization is halted when the vessel is depleted of monomer, but the chain ends are still active—”living polymers.” (Id.)
McGHEST/Polymerization suggests use of peroxides, persulfates, azo compounds, oxygen and certain radiation for free radical initiation; Lewis acids (boron trifluoride), sulfuric acid, aluminum chloride or Friedel-Crafts reagents as cationic initiators, and chromium oxide (on silica-alumina), nickel, cobalt (on carbon black), molybdenum (on alumina), alkylaluminum, and titanium chlorides, in hydrocarbon or ether solvents, for anionic initiation.
Catalysts participate in a reaction, speeding it up, but aren’t consumed by it. If the NTL chemists are unhappy with polymerization times, they will try adding catalysts. If they’re lucky, Grantville literature will suggest a specific catalyst for the polymer of interest. Otherwise, they will run through their “medicine chest” of common catalysts—typically strong acids and bases, metals, and metal oxides—and see what works.
The right catalyst can also direct the monomers to connect in a very specific way, changing the properties of the polymer. (See “Stereochemical Control Problems”, below).
In table 5-6, I propose a timeline that covers some of the more important of the substances that can be used as polymerization initiators or catalysts, or as catalysts in monomer synthesis, or as supports for catalysts.
Table 5-6: Initiator/Catalyst/Catalyst Support Timeline
sulfuric acid, nitric acid, hydrochloric acid, potassium and sodium hydroxides, ammonia, copper, silver
carbon black, cuprous chloride, mercuric sulfate
oxygen, zinc acetate
nickel, sodium, hydrogen peroxide, metal peroxides, phosphorus trichloride, hydrogen iodide, hydrogen fluoride
aluminum oxide (alumina), chloride and iodide, chromium oxide and chromic acid, platinum, peroxydisulfates
cobalt, molybdenum, alkylaluminum, lithium chloride
titanium chlorides, palladium, rhodium, vanadium oxide, vanadium, lithium, alkyllithium
potassium, Ziegler-Natta and metallocene catalysts
Note that I have moved lithium one row earlier on the timeline from where it was in Chapter 2, because of the demand for lithium chloride and alkyllithium as polymerization catalysts.
Polymerization Control Problems
At one extreme, we can have trouble achieving a reasonable degree of polymerization. Low-density polyethylene (LDPE) may be made with just a peroxide or peroxysulfate catalyst, but the pressures needed were among the highest used in the Forties chemical industry. This in turn resulted in high plant investment costs, about $1000/ton annual capacity. And the product itself had problems; it was too soft, weak and flexible, and couldn’t be sterilized. (Sittig 12ff).
Paraffin wax (under 1000 molecular weight) is essentially what you would get if you tried to make polyethylene and failed to achieve a high enough degree of polymerization.
The most effective catalyst for polymerizing formaldehyde (to make polyacetal resin) is triphenyl phosphine. CCD 891 says that this is made by a “modified Grignard synthesis,” which is not too informative. Fortunately, it’s possible, although apparently more difficult, to just use a strong acid. Hence, I deemed that the catalyst was not a limiting factor.
Polyacetal depolymerizes at 110–120oC and to prevent this, it must be “capped” by reaction with acetic anhydride or suitable co-monomers incorporated. (EB15). Either acetyl chloride or acetic anhydride is used as the acetylating reagent in the manufacture of aspirin and since aspirin is in canon (as of August 1631!—see Offord, “Dr. Phil’s Amazing Lightning Crystal”, Grantville Gazette 11) we may have figured out how to make acetic anhydride at an early date. The recommended co-monomers are ethylene oxide or 1,3-dioxane. (McGHEST/polyacetal).
The Germans relied on methyl rubber—a polymer of dimethylbutadiene—during WW I. Polymerization times were 6–10 weeks to make H-rubber (hard) and three to six months to make W-rubber (elastic). (PSLC) The original process for making styrene-butadiene rubber (Buna S) took weeks to convert the monomers, but after years of development work this was reduced to hours (Miles 302).
It can also be difficult to drive a condensation polymerization to completion. Grantville chemists will be able to write the chemical reaction for these polymerizations and in theory should recognize that according to Le Chatelier’s principle, removal of the byproduct will drive the reaction forward. But bear in mind that Carothers, who won a Nobel Prize for his work on polymers, didn’t anticipate that the polymerization reaction of a dialcohol and a diester, to make the first synthetic heteropolymer, would be inhibited by water production. (Meikle 130). And it took him about a year to get a product with a molecular weight above 5000. (Teergarden 61).
Still, EB15 says “in the case of condensation reactions, reactors must provide for the efficient removal of volatile by-products.” In a condensation reaction, some small molecule is produced by each addition step. If the concentration of this byproduct increases, it will inhibit the polymerization. It can be eliminated physically (escaping as a gas if volatile, escaping by precipitation if insoluble) or chemically (by reaction with something else).
For example, the nylon 6,6 polymerization is a condensation with a water byproduct, so water causes depolymerization. A “molecular still” was needed to more completely eliminate the water and thereby drive polymerization to completion. (Teergarden 126).The reaction of bisphenol A and phosgene to produce polycarbonate releases HCl; this can be destroyed by conducting the reaction in a basic medium. (Kirk Othmer/Polymers 19:897).
At the other extreme, the degree of polymerization can be too great. For chromium oxide-mediated polymerization of ethylene, it’s necessary to include “chain-stoppers”, or one obtains a very high molecular weight polymer that’s impossible to process further. (Wittcoff I-199). And I am afraid that the Grantville researchers are going to run afoul of this problem and have no inkling of what’s going wrong, let alone how to fix it, because they have so little real knowledge of polymerization processes and little in the way of analytic methods for characterizing the polymer.
Similar problems exist for the manufacture of nitrile (acetonitrile-butadiene) rubber. The Grantville chemists haven’t been warned that if they don’t add a “chain stopper,” the polymerization will create a very high molecular weight, highly branched polymer that is essentially impossible to process. (Kirk-Othmer, 8:1011).
With regard to styrene-butadiene rubber, CCD 822 refers to the value of dodecyl (lauryl) mercaptan as a chain-modifying agent. That chain-modifying agent is actually critical to obtaining a processible polymer. (Kirk-Othmer 8:914). I have not found a synthesis description in Grantville literature.
EB15 describes two methods for making Bakelite: (1) react phenol with excess formaldehyde, catalyzing with base to obtain a “resole” (prepolymer, “A-stage”), then heat to obtain thermoset network polymer; and (2) react formaldehyde with excess phenol, catalyzing with acid to obtain “novolac” (another prepolymer), then supply more formaldehyde. We may for example use aqueous 37–50% formaldehyde at 50–100oC (CCD 669) with an ammonia catalyst (EB15/Ammonia).
I brooded for quite a while over whether we could make Bakelite (phenol-formaldehyde resin) any earlier than its first canon appearance (1634, see Offord, “Feng Shui for the Soul,” Grantville Gazette 17).
On the one hand, phenol can be distilled from coal tar and formaldehyde produced by oxidation of methanol (itself from destructive distillation of wood). We have DDT in June 1633 and so we had to have chlorobenzene earlier. The benzene would have had to have been extracted from coal tar. So we probably had both monomers in hand by late 1632.
On the other hand, the OTL development of Bakelite was marred by false starts. An 1891 attempt to react phenol and formaldehyde resulted in “violent release of gas at the beginning of the reaction,” producing a porous, brittle material. Still, its hardness and chemical stability were attractive. Most further efforts concentrated on finding a plasticizer, analogous to camphor for celluloid. In 1907, Baekeland chose instead to use heat (150oC) and pressure (100 psi), with various catalysts (HCl, zinc chloride, ammonia), to suppress the foaming. It was somewhat analogous to processes already used to harden shellac and vulcanize rubber. (Meikle 36ff).
Baekeland’s patent (939966) explained that the foaming occurred when the initial condensation product (the A-stage) was heated in an open vessel above 100oC. If a lower temperature was used, the final hardening was very slow and possible only with thin material. Baekeland taught to instead use counter-pressure. (Nor was this disclosure the end of the R&D process; Baekeland found that he had to offer considerable technical assistance to customers to ensure that the plastics were properly formulated and processed to do the job. (Meikle 41ff)).
The question, then, is how much of the Baekeland “secret” was revealed by Grantville literature. We are told that “high temperature and pressure” were needed (EB15/Baekeland), and perhaps that’s enough.
Butyl Rubber. Isobutylene was first polymerized in 1873, with the goal of making a substitute for natural rubber. Unfortunately, the polymer couldn’t be crosslinked to form a useful elastomer. (Kirk-Othmer 8:934) The solution was incorporation of a small amount of isoprene. Even so, “Before experimental difficulties were resolved, butyl rubber was called ‘futile butyl'”.
I would be more sanguine about the prospects for making butyl rubber (admittedly much inferior to natural rubber) if we knew what those “experimental difficulties” were.
Stereochemical Control Problems
The properties of some polymers is dependent, not only on having the correct monomer, but having those monomers correctly oriented along the polymer chain. It’s usually necessary to use a special initiator or catalyst to obtain the desired sterochemical control, and those catalysts may be hard to come by. A “plain vanilla” catalyst like peroxide or peroxysulfate might not work at all, or it might yield a polymer with a disfavored stereochemistry. Table 5-7 reviews the “problem” catalysts and polymers.
Table 5-7: Selected Polymerization Requirements
styrene-butadiene rubber (SBR)
I will discuss the catalysts first, then comment on some of the individual polymers.
Chromium trioxide is likely to be the first HDPE catalyst available in the 1632verse. There is a De Geer-sponsored, Richelieu-approved expedition to Maryland to seek chromite (iron magnesium chromium oxide) for making stainless steel; if they are very lucky they will start production by 1635. Chromium trioxide may be made from potassium bichromate (EB11/Chromium), the latter being obtained by “fusing chrome ironstone with soda ash and lime” with a secondary treatment with sulfuric acid (EB11/chromates and dichromates). The trioxide should not be confused with the sesquioxide produced as green “ash” in the classic ammonium dichromate “volcano” experiment.
Ziegler-Natta catalysts are a combination of a transition metal halide (e.g., titanium tetrachoride) and organometallic compound (e.g., alkylaluminum). Grantville literature provides some information about them (See Appendix for particulars). We can at least sketch out a process for making the two components. Titanium tetrachloride is made by heating titanium dioxide in chlorine. (CCD866). Titanium dioxide in turn may be obtained from certain sands in India. (see Cooper, “Gajam Raanni,” Grantville Gazette 25). Triethylaluminum is derived by reacting ethylene, hydrogen and aluminum under “moderate temperature and varying pressures” (CCD881).
What the Grantville literature does not say is just how devilish it will be for a fledgling chemical industry to make and use these complex, finicky catalysts. There is no disclosure of the ratio of the components or the size of the particles, or how the components should be ground together.
Ziegler-Natta catalysts are decomposed by oxygen, carbon dioxide, water, sulfur, acids, alcohols, halogens, amines and other substances. In 1961, ethylene was commercially available in a form containing not more than 1500 parts per million oxygen—and that’s not good enough! You want to get it down to 50, 10 or even 5 ppm. (Sittig 29ff, 115). Kirk-Othmer/Polymers (19:902) observes that these catalysts are “easily poisoned by moisture.”
The simplest way of getting the reactive impurities out of the monomers is to pre-treat them with the same or similar organometallics until there’s none left to react. Of course, that’s wasteful of a very expensive organometallic, unless it can be recovered efficiently.
Another problem with the catalysts is that they are pyrophoric (Id.). Given that the monomers (and probably also the solvent) are flammable, this means that they are risky to use given the rather crude state of industrial safety in the 1632verse.
If these problems are overcome, the experimenters will still need to work out the precise temperatures, pressures and process times. There’s also the need to recover the catalyst, not merely for economic reasons, but because if left in the polymer it impairs its properties and also corrodes the polymer molding equipment. (Sittig 119).
Metallocenes are organometallic compounds in which a metal atom is sandwiched between two rings. Metallocene polymerization catalysts became commercially prominent in 1995, and hence they are likely to be mentioned only in EB15 versions published after that date. Since the McGraw-Hill Encyclopedia of Science and Technology in the high school library is the 1977 edition, it’s highly unlikely to say anything about it. However, I discuss the teachings of the 2002 edition in the Appendix.
Alkyl Lithium. Butyl lithium is manufactured by reacting lithium ribbon with a butyl chloride in pentane or hexane. (CCD 139). Lithium metal is not one of the easier ones for us to obtain (see Cooper “Mineral Mastery” and “Industrial Alchemy, part 2”), but of course we don’t need huge quantities just to make a catalyst.
Peroxides. Peroxide manufacture is discussed in Cooper, “Industrial Alchemy, Part 2: Inorganic Chemical Bestiary” (Grantville Gazette 25).
Now, a few comments on the problem polymers . . . .
Polypropylene. Produced like LDPE, that is, at high pressure, polypropylene is atactic, and essentially worthless. Chromium oxide is not useful in polymerization of propylene. (EB15/Transition Element). The Ziegler-Natta catalysts make possible the production of isotactic and syndiotactic polypropylene. (Solomons 409ff). However, it should be noted that while Ziegler used the titanium tetrachloride in the catalyst system for HDPE, Natta found it necessary to use the trichloride in producing stereoregular polypropylene. (Miles).
Ethylene-Propylene Monomer rubber (EPM) is a block polymer of ethylene and propylene; EPDM adds a small amount of a diene. According to EB15, both are made using Ziegler-Natta catalysis. The same is true of the fluorinated analogue of EPM.
Synthetic Rubber, Generally. Natural rubber is a polymer of cis-isoprene. Attempts to synthesize natural rubber replacements (whether from isoprene or butadiene) resulted in polymers which were random combinations of cis- and trans- structures, until first, it was recognized that natural rubber structure was stereospecific (the “cis” means that the hydrogens of the C=C carbons are on the same side of the double bond), and second, stereospecific polymerization catalysts were developed. That didn’t happen until the mid-Fifties.
Polybutadiene Rubber. The first polybutadiene was made after World War I using sodium (or other alkali metal) catalyst. This was the original BuNa rubber (Na=sodium), and it was 10–23% cis-1,4, 25–45% trans-1,4, and 23–65% “side vinyl” (Morton, 247). Obtaining sodium metal is not a problem. The problem was that it was quite inferior to natural rubber.
A 95–97% cis-1,4 polybutadiene, needed for tire rubber, is in fact made with Ziegler-Natta catalysts, or catalyst systems of equal complexity.
A 35% cis, 55% trans, 10% “side vinyl” polybutadiene is made with a “living anionic catalyst” based on alkyl-lithium, and lends itself to creation of modified polybutadienes. Similar catalyst systems can be used to prepare medium or high vinyl polybutadienes. The high vinyl polybutadienes also vary in terms of stereoregularity (atactic, isotactic, syndiotactic). There is also a high-trans polybutadiene similar to gutta-percha that EB15 ignores. Like the cis-rich polymer, it requires transition metal catalysis. (Kirk-Othmer 8:1031ff, IISRP; Yoshioka).
Polyisoprene Rubber. Unfortunately, the more available alkyllithium catalysts do not elicit as high a cis-content as do Ziegler-Natta types. The polymerization is also acutely sensitive to the structure of the Ziegler-Natta catalyst. Only if the catalyst contains beta-titanium trichloride do you get high-cis polymer; the alpha, delta and gamma forms yield trans. It’s possible to make a polyisoprene by free radical (persulfate) emulsion polymerization. Unfortunately, you obtain a high-trans structure, resembling gutta-percha. (Kirk-Othmer 8:915, 9:4).
Styrene-Butadiene Rubber. EB15 teaches free radical initiation in emulsion, or anionic initiation in solution, resulting in a random copolymer. CCD822 says to use a 1:3 ratio, and that a peroxide catalyst is fine.
Historically, SBR went through four phrases. First, there was the “hot” emulsion polymerization using free radicals (potassium peroxydisulfate), resulting in the “Buna S” of the Thirties. The butadiene units were about 72% trans isomer. Second, during WW II, the Germans developed “cold” methods, with a “redox” initiator, which were adopted by American companies after the war. The “cold” polymer was more linear. Third, in the Fifties, the Ziegler-Natta catalysts made possible SBR with a high-cis isomer content. Finally, in the Sixties, we switched to anionic solution polymerization with organolithium catalysts, which resulted in a “living” polymer suitable for grafting. (Kirk-Othmer 22:995).
We can certainly make the original Buna S (once we have styrene and butadiene), and somewhat later the “living” SBR, but the high-cis SBR may be more of a problem.
Polymethylene. In view of the potential problems with early duplication of the historical processes of making polyethylene, I have investigated the prospects for use of polymethylene (a linear molecule like HDPE, except that it can have an odd number of carbon atoms) as a substitute. Polymethylene was made by Von Pechmann (1898–9), by thermal decomposition of diazomethane in ether (EB15).
Diazomethane may be made by reaction of chloroform, hydrazine and potassium hydroxide, or of nitrosomethyl urea with potassium hydroxide. (MI342; cp. EB11/Diazo Compounds).
There was no attempt to commercially exploit polymethylene produced by thermal decomposition. Diazomethane was a rather expensive material (Lokensgard 460), and dangerously explosive, and hence even if the modern uses of polyethylene were foreseen it probably wouldn’t have been commercially feasible to use it.
In 1948–52 it was discovered that the decomposition and polymerization could be catalyzed by alkyl boron esters or by copper powder. Unfortunately, the resulting polymers were brittle, of low molecular weight (18–22K), and obtained in very low yield. Rawe (267) says that “formation of polymethylene by this reaction is not practical for commercial utilization.”
A 1956 patent (US 2749318) describes obtaining a more useful polymer using, as catalyst, boron trifluoride, and a 1962 patent (3062756), using an alkali metal borohydride or a boron-containing carborane. The ‘318 patent said that the product had a molecular weight of 3300K, tensile strength of 4900 psi, and extensibility of 500%—better than LDPE—and good electrical properties. However, there are no signposts to direct Grantville chemists to these catalysts, and in any event, it doesn’t appear that these processes were ever shown to be commercially feasible.
Heteropolymer Structure Control
In a heteropolymer, there are different ways of arranging the monomers. Think of a pearl necklace as a polymer, with black and white pearls representing two different units. They can alternate, they can be grouped in blocks of the same color, or they can be randomly strung on the necklace. Those are analogous to alternating, block, and random copolymers, respectively. Next, imagine that the main string is all black pearls, but there are little strings with white pearls hanging down from it. That’s analogous to a graft copolymer.
An alternating copolymer results automatically if one of the comonomers is unable to react with itself, or if the two comonomers form a complex that is what actually is polymerized.
Otherwise, the most likely product is a random copolymer, but the proportions of the monomeric units will depend on the original monomer proportions and their relative rates of polymerization. There are many useful copolymers that are predominantly derived from one monomer, with just a smidgeon of the other.
Special methods are needed to make block and graft copolymers. Spandex is an example of a block copolymer, and impact polystyrene or ABS of a graft copolymer. (Wittcoff I:192).
You can make a block copolymer by first making a living homopolymer (a polymer that still has an active chain end) of monomer #1, then adding monomer #2, then adding monomer #1 again, and so on. EB15 mentions “living polymers” but doesn’t provide details about how to avoid chain termination. Living polymers are in fact the natural result of anionic polymerization because you have to add a “chain stopper” to terminate the polymerization. (Teegarden 99).
Graft copolymers are made by creating free radical sites along the backbone, to initiate the growth of secondary chains, or by providing free reactive groups along the backbone that a second monomer can be condensed onto. (Wittcoff I-192).
ABS (acrylonitrile-butadiene-styrene) is a graft polymer. The earliest forms of ABS were either mechanical mixtures of styrene-acrylonitrile and butadiene-acrylonitrile (nitrile rubber) polymers, or obtained by co-coagulation of the two polymers. The modern ABS is obtained by grafting acrylonitrile and styrene monomers onto styrene-butadiene or polybutadiene rubber. That is, the styrene and acrylonitrile are allowed to polymerize in an emulsion formed by the latex. (Scheirs 19).
Hence, to make ABS, we need to have at least one of the three elastomers: nitrile rubber, high-cis SBR, or high-cis polybutadiene.
I wonder whether some bright chemist will think of grafting styrene and acryonitrile onto natural rubber, which is poly-cis-isoprene, thus obtaining semisynthetic AIS (acrylonitrile-isoprene-styrene). (Prasassarakich).
In part B, I have defended my prediction that certain polymers will appear relatively late in the 1632verse. If you think I haven’t been pessimistic enough, you should look at the Appendix (which will be posted to www.1632.org ), as this contains additional detailed analysis of what Grantville literature has to say about synthetic routes to the various monomers, and polymerization conditions.
The band Aqua had “Barbie Doll” sing, “Life in plastic, it’s fantastic.” Hopefully, the down-timers will soon be singing, “Life with plastic, it’s fantastic.”