Home > News > Content

PCBs: Products And By-Products

Between 1929 and 1989 an estimated 1.7 million metric tons of PCBs were produced commercially worldwide.2 These commercial PCBs were sold in the United States under dozens of different trade names, of which the most commonly known was Aroclor. These products were mixtures that contained different PCB congeners, each with between 1 and 10 chlorine atoms.3 (Unintentionally produced PCBs also typically occur as mixtures of congeners.) Each of the total 209 PCB congeners has a distinct structure that influences how it behaves environmentally and how it interacts with living cells and organisms.
Although large amounts of commercially produced PCBs remain in the environment, monitoring conducted at a number of sites around North America in the past five years or so has detected notable air concentrations of PCB congeners that are not part of the historical, now-banned commercial PCB products. These congeners have also been detected in U.S. water bodies including the Delaware River, the New York–New Jersey Harbor, the Houston Shipping Channel, and San Francisco Bay. In the Spokane River in Washington State, such PCBs are currently responsible for violating local water quality standards.4
One of these congeners, PCB 11, is emerging as a marker of nonlegacy PCB contamination. In what amounts to environ-mental and scientific detective work, researchers who detected PCB 11 and other congeners in air samples discovered that these PCBs matched those previously detected in water discharged from paint manufacturing facilities. Testing of paint samples has turned up 50 different PCB congeners, including PCB 11, which is known to be produced in the manufacture of diarylide yellow pigments. PCB 11 is neither associated with historical commercial PCB products nor a breakdown or a dechlorination product of those old commercial mixtures.5 So if PCB 11 is being found in the environment, it is most likely coming from sources other than legacy PCBs.6
PCB 11 has been found in Great Lakes sediment7 and in polar region air samples.8 Work by a number of research teams points to paints and pigments as likely sources of the PCB 11 being detected environ-mentally.7 But PCB 11 is far from the only congener being found in testing of pigments and the products they are used in. Such testing has revealed PCBs ranging from those with just a few chlorine atoms, PCBs 1 through 11, to the most chlorinated congeners, PCBs 206 through 209. These congeners include a number of those previously identified as dioxin-like.11 Dioxin-like PCBs are structurally similar to polychlorinated dibenzo-p-dioxins and dibenzofurans, chemicals known to be environmentally persistent and to have adverse health effects, among them cancer and impacts on development, reproduction, and immunity. Biologically, these chemicals share an ability to bind to a cellular receptor known as the aryl hydrocarbon receptor, which regulates the expression of genes that influence multiple functions, including enzyme production and hormone regulation.
Recent testing of pigments in Japan by the Japan Dyestuff and Industrial Chemicals Association has found traces of PCBs in 57 out of 98 organic pigments tested.9,10 Some of these pigments were found to contain PCBs at concentrations above 50 ppm. Testing of newspaper and magazine papers, food packaging, and plastic bags colored or printed with inks and pigments associated with PCB by-products has also confirmed the presence of these PCBs, including PCB 11.5
The PCBs detected in paint and pigment samples and those detected in finished products colored with such pigments also overlap with those identified in 2012 testing of effluent from the Inland Empire Paper (IEP) Company recycling facility in Spokane, according to IEP environmental manager Doug Krapas. IEP was able to connect the PCBs in its wastewater with inks on the paper it recycles, much of it newspapers, magazines, mailing materials, and packaging. Papermaking processes historically have been a source of chlorine compounds that are known to have adverse health effects on aquatic ecosystems, but over time IEP had changed its processes to eliminate the use of chlorine. It had also installed secondary and tertiary treatment and closed-loop systems to reduce overall water use and water discharges and to reduce contaminants in its wastewater as much as technically possible. When the company discovered that, despite all of these measures, PCBs exceeding local standards were being found it its wastewater, it realized that something in the paper being recycled must be causing the problem.
According to Krapas, as long as recycling facilities continue to process paper that contains ink with allowable levels of by-product PCBs, there will be problems meeting water quality standards. Inkborne PCBs present a particular problem for IEP’s Spokane plant, given the low levels of PCBs allowed by local water quality standards—local Spokane Tribe water quality standards limit PCBs to 3.37 parts per quadrillion (ppq).12 This standard is based on fish consumption—that is, the amount of PCBs people would ingest by eating fish from the Spokane River—and is based on the tribe’s reliance on local fish as a staple food.
Keri Hornbuckle, a University of Iowa professor of civil and environmental engineering, says that prior to 2007 most environmental monitoring for PCBs was not designed to sample for all 209 PCB congeners but focused only on those that were produced intentionally, the commercial Aroclor mixtures. This means that environmental monitoring was likely missing detection of PCBs that are unintentional by-products. New monitoring technology used by Hornbuckle and colleagues through funding from the Superfund Research Program of the National Institute of Environmental Health Sciences—initially to monitor air in Chicago but since conducted in many other locations—was what for the first time profiled the environmental presence of unintentionally created PCBs.
Hornbuckle and her colleague Dingfei Hu, a University of Iowa assistant hydroscience and engineering research scientist, hypothesized that because the PCB congeners they were detecting in air samples were the same as those that had been detected in water discharged from paint manufacturing facilities, these PCB congeners might also be present in commercial paints. To investigate, the researchers measured PCBs in paint pigments purchased on the U.S. retail market in 2009.6 Their analysis found more than 50 different PCB congeners (including several previously identified as dioxin-like) in the pigments. The identified PCB congeners appeared to vary depending on the types of pigments analyzed and the manufacturing processes involved in their production. Although the PCBs generated as by-products vary by pigment type, the processes that result in these PCBs typically combine chlorine, salts, and hydrocarbons or chlorinated hydrocarbon compounds at high temperatures.
Further analysis indicated that certain PCBs were prevalent in what are called azo, diarylide, and phthalocyanine pigments, which are commonly used to color inks, dyes, paint, paper, textiles, plastics, leather, cosmetics, and foods, among other materials and products. Azo and diarylide pigments are used primarily to make yellows but also some reds and oranges, while phthalocyanine pigments are used primarily to make blues and greens.
“The widespread use of these pigments explains the presence of PCB 11 in commercial goods common throughout modern society, such as newspapers, magazines, and cardboard boxes,” wrote Hu and Horn-buckle. “Although we do not know if inadvertent PCBs have adverse effects on human health, there are many potential routes for human exposure to these PCBs through inhalation, dermal exposure, and ingestion due to their physicochemical characteristics of semi-volatility, hydrophobicity, and persistence.” They also say that to their knowledge, “pigments or dyes are the only significant source of PCB 11,” and therefore detection of PCB 11 in air “must be associated with human activity utilizing pigments or dyes.”6
In a subsequent published study, Hu and Hornbuckle noted that the pattern of PCBs detected in the commercial paint pigments tested was unrelated to that of commercial PCB production and included “many congeners that are highly bioaccumulative, dioxin-like and/or probably carcinogens.”7 They cited paint production and use as the probable source of these PCBs in North America. Their investigation, which focused on Great Lakes sediment, appeared to confirm the presence in the environment of manufacturing by-product PCBs not associated with historical commercial mixtures. They do note, however, that because their sediment study did not enable them to identify sources of PCB congeners that appear to be common to both legacy commercial Aroclor mixtures and pigment by-products, it leaves important questions unanswered about dioxin-like PCB congeners found in these samples, which have potential multiple sources,
Further establishing the link between pigments and PCBs in the environment is research by Lisa Rodenburg, a Rutgers University associate professor of environmental science, in which she directly tested products printed or colored with pigments associated with by-product PCBs.5,13 Among the printed products Rodenburg tested were colored newspapers, glossy magazine papers, plastic bags, and cereal and other food boxes. The PCBs she identified in these printed paper and plastic samples, primarily those colored yellow, were the same nonlegacy congeners found in environmental samples and in paint pigments by Hu and Hornbuckle. They also matched the PCBs being measured in effluent from IEP’s Spokane recycling facility.
In addition to organic pigments, Rodenburg’s research indicates that certain titanium dioxide manufacturing processes can also produce PCB by-products.13 Although Hu and Hornbuckle did not find PCBs in their testing of inorganic pigments, including those containing titanium dioxide, titanium dioxide manufacture has been cited as a source of by-product PCBs by the Michigan Department of Environmental Quality,14 the New York Academy of Sciences,15 and the Australian government,16 among others.
Hu and Hornbuckle’s research suggested several possible ways in which by-product PCBs are created during manufacturing. The phthalocyanine blues and greens analyzed so far have tended to contain higher-chlorinated PCB congeners (those with more chlorines) than the diarylide yellows tested. As described by Hu and Hornbuckle, the process commonly used to produce these blues and greens involves phthalic anhydride, urea, or phthalonitrile and a copper or copper salt, which are then processed using an organochlorine solvent such as di- or trichlorobenzene.6 As noted above, the combination of chlorinated hydrocarbons and salts processed at certain high temperatures is what leads to the creation of PCBs.
Hornbuckle also explains that the highly chlorinated PCB by-products in blue and green pigments behave differently than the lighter-weight, less chlorinated by-products in the yellow pigments. The lighter, less chlorinated chemicals are more volatile and therefore more mobile in air and can be expected to move out of the pigmented paints and inks more readily than heavier, more chlorinated chemicals would. The heavier blue and green pigment by-products are likely to stay with paints when they chip, thus enabling the by-product PCBs to transfer to soil if these paints are used outdoors—on building exteriors, for example.
According to research by Delaware River Basin Commission geologist Gregory Cavallo along with Rodenburg and colleagues at Rutgers University, about 65% of all organic pigments produced go into printing inks (as opposed to dyes, paints, or other colored products). In 2006 one-quarter of the worldwide production of 250 million metric tons of organic pigments was estimated to be diarylide yellow pigments. Analysis of printing inks indicates that the typical printing ink contains an estimated 40% pigment. Cavallo et al. therefore estimate that worldwide production of PCB 11 through the manufacture of diarylide yellows would have been about 1.5 metric tons in 2006.5