From a professional perspective, PCBs entered my life in 1978 while I was post-grad research associate at the U of Hawaii College of Tropical Agriculture.  The mission of our lab was to develop data in support of EPA pesticide registrations for tropical crops.  Pesticide registration is normally conducted by the pesticide manufacturers, but tropical crops are such a small niche market, that it isn’t worth their trouble in most cases.

One day the lab director brought a box with ten 8-ounce jars into my work area and put them down on my lab bench (tangential comment – this is the work area from which I had a view of three waterfalls, sigh).  He told me each of the jars contained a different Aroclor PCB mixture and that the Hawaiian electric company wanted us to develop an analytical method to measure the amount of PCBs in transformer mineral oil.  For the next couple of months, while working up the analytical method, these jars were front and center on my lab bench.  These were not laboratory prepared analytical standards; these were jars containing pure (“neat” to you chemists) Aroclors.

As a young chemist the opportunity to work on an environmentally relevant project was a real thrill.  As you likely know, the percent of chlorine in an Aroclor is indicated by the last 2 digits in the model number; so Aroclor 1221 contains 21% chlorine by mass and Aroclor 1268 contains 68% chlorine (Note: this numbering does not apply to Aroclor 1016 which is a modified version of Aroclor 1242 and thus contains 42% chlorine).  The lighter Aroclors like 1221 and 1232 were as thin as machine oil.  The mid weight Aroclors (1248 and 1254) were as viscous as motor oil.  Aroclor 1268 was pretty much a solid at room temperature.  Also, the lighter Aroclors were clear and the heavier ones had a darker quality to them.

On the chemistry side, there are 209 different PCB molecules (called “congeners”), and each of the Aroclors is a mixture of 50 or more of these congeners.  Chemists sometimes organize the different congeners into groups based on the number of chlorine atoms they have and these groupings are called “homologs”.  So for instance, all the different congeners that have three chlorines belong to the tri-chloro homolog group, all the congeners with four chlorines belong to the tetra-chloro homolog group, and so on.

There is an interesting difference among the Aroclors (interesting to me at least) that even many chemists are not aware of; each Aroclor PCB mixture is dominated by a different homolog group.  So for instance, Aroclor 1221 is made up of 60% mono-chloro congeners (one chlorine), Aroclor 1248 is 56% tetra-chloro congeners (four chlorines), and Aroclor 1262 is 47% hepta-chloro congeners (seven chlorines).  This formulation was not created by design; it was just an accident of the manufacturing process.

Meanwhile back at my lab at the U of H, and many, many gas chromatographic runs later, I did finally come up with a reliable method for measuring PCBs in transformer oil.  Remember this was before there was an SW-846 (EPA’s compendium of analytical methods) or a Method 8082 (EPA’s PCB analytical method).  The method I developed for measuring PCBs in transformer oils was actually published in the Journal of Chromatography (JOC) and the article can still be obtained on-line (the link is for the curious, but purchase is not recommended since this method has been superseded by better USEPA analytical procedures).

When I looked up the link to the JOC article to include with this post I was disappointed to see that the U of H College of Tropical Agriculture now has a trendier name, the “Department of Agricultural Biochemistry”.  On the other hand, I was very pleased that when I clicked my name on the author list my address came up as “University of Hawaii, 1800 East-West Road, Honolulu, Hawaii”;  on a cold windy late winter day it’s nice to still have an address in Hawaii.

In closing, let me acknowledge Dr. James Ogata who directed my work at the U of H lab and who prepared the manuscript for publication.

For help with PCB chemistry questions, please contact me at okun@oto-env.com.


Aroclor 1254 was the PCB mixture most commonly used in building materials, based on my personal experience reviewing test results.  Recently, as I was digesting data and researching literature to prepare for my UMass Soils Conference presentation this fall, I discovered something interesting about Aroclor 1254 that I hadn’t known; there are two types of Aroclor 1254, and while they are similar in their physical properties, from the standpoint of environmental toxicology they are very different.

Type 1 Aroclor 1254 was one of the commercial PCB mixtures manufactured and sold by the Monsanto Company prior to 1971; a high percentage of its production was used in electrical equipment like transformers and capacitors, smaller amounts were used in other applications like being formulated into building materials (e.g. paint and caulk).  In its pure form Aroclor 1254 (Type 1 and Type 2) is an highly viscous liquid that is thicker than honey.  While each batch of PCBs differed slightly in its exact chemical composition (based on the percentages of the the individual PCB congeners present), these differences from batch to batch were small. However, around 1971 the formula for making Aroclor 1254 changed in a fundamental way giving rise to what is now referred to as Type 2 Aroclor 1254.

Type 2 Aroclor 1254 was an indirect result of the 1968 Yusho rice oil poisonings. After the Yusho incident, Monsanto (the sole US PCB manufacturer) was seeking alternatives to reduce the toxicity of its PCB mixtures.  They hit on two ideas:

  1. Introducing a new PCB mixture, called Aroclor 1016, that contained lower concentrations of the volatile low molecular weight congeners and the more toxic high molecular weight congeners; and
  2. Stopping the sale of PCBs for use in applications that were not considered to be “totally enclosed”.  The totally enclosed requirement was intended to limit PCB uses that were more likely to result in human or environmental exposures.

Aroclor 1016 was produced by first making Aroclor 1242 (which it resembles), and then through distillation removing the light and heavy ends of the mixture.  As predicted, the toxicity of Aroclor 1016 was significantly less than that of other PCB mixtures; so far so good.   However, at this point in the story you might want to ask: “what happened to the light and heavy ends removed from the Aroclor 1242 to make Aroclor 1016?  Were these simply discarded?”

The answer is no, they were not discarded.  These light and heavy ends were used as feedstock for the manufacture of Type 2 Aroclor 1254.  Type 1 Aroclor 1254 was manufactured as a one step process; biphenyl (derived from coal tar) was chlorinated until the chlorine content in the resulting PCB mixture reached 54%.  In contrast Type 2 Aroclor 1254 was manufactured using a two step process where the by-products of Aroclor 1016 manufacturing (the light and heavy ends) were re-chlorinated to end up with a final product with 54% chlorine.

The physical properties of Type 1 and Type 2 Aroclor 1254 are virtually indistinguishable, but their chemical, toxicological and likely environmental properties have been shown to be different.  Type 2 contains higher concentrations of the so-called “co-planer” congeners (for chemists: that means the ones with no ortho substitutions) including about 5 times more of the three most toxic of these co-planer congeners, PCB 126, PCB 169 and PCB 77.  Type 2 Aroclor 1254 has also been reported to contain significant concentrations of the highly toxic polychlorinated dibenzofurans (PCDFs); these are virtually absent in Type 1 Aroclor 1254.

One interesting question is how much of the total Arochlor 1254 production was Type 2?  The answer from Monsanto’s production records indicates that about 1% of total Aroclor 1254 was Type 2.  Since Monsanto stopped PCB sales for other than totally enclosed uses at about the same time that production of Type 2 Aroclor 1254 began, it is likely that most or all of it went into electrical equipment like transformers and capacitors with little if any going into building materials.

Another interesting observation is that it is likely that Type 2 Aroclor 1254 is the PCB mixture that has been most commonly used in conducting toxicological studies on PCBs.  This is because towards the end of Monsanto’s PCB production (after 1970), most or all of the Aroclor 1254 being produced was Type 2.  As a result, this was the Aroclor 1254 that was available for distribution to researchers when toxicological studies were being undertaken on PCBs.  It took almost 20 years after the end of US PCB production before scientists could detect the chemical differences between Type 1 and Type 2 Aroclor 1254.

For more information please email me at okun@oto-env.com.