Secret Knowledge
Larry Ytuarte, Ph.D.

November, 1991.

    Medical examiner Robert Bux, of the Bexar County Forensic Science Center (which back in 1991 was called - take a deep breath - the Bexar County Medical Examiner’s Office and Regional Criminal Investigation Laboratory) in San Antonio, Texas, requested an arson analysis.  The analysis was to be performed on remnants of clothing removed, during autopsy, from the body of an elderly man who had died in a house fire in Fort Stockton, Texas.  The purpose of the analysis was to determine if gasoline had been used to start the fire.

    The deceased’s name was Bill Richardson.  Richardson had a niece, Sonia Cacy, who was living in the house with him at the time.  On the night of the fire, Cacy was able to get out of the house, but her uncle died inside.

    Sometime after the fire had been put out, while she was still at a hospital where she was recovering from smoke inhalation and shock, Cacy was visited by police officers.  They brought a warrant.  They left with scrapings from her fingernails and samples of her blood.  Arson was suspected.  And Cacy was the suspect.

Sonia Cacy mugshot

    So, how did Cacy so quickly become the suspect?  Well, her uncle had made a hand-written will naming her as his beneficiary.  This will had been found in her bedroom, a bedroom in the very house that had burned.  (This might be a good time to mention that Bill Richardson didn’t have any insurance.  And as for the house: Who burns down his or her inheritance in order to inherit it?  And who leaves a will in a house that he or she is deliberately burning down if that same will is going to be the ticket to an inheritance?)

    I was working as a toxicologist in the lab at the time.  (I’d begun working at the lab in September of 1990.  I would be fired in September of 1994 after blowing the whistle on a variety of criminal and unethical practices in the lab.)  Assistant Chief Toxicologist Joe Castorena explained to me why Sonia Cacy had murdered her uncle by piling furniture on top of him, pouring gasoline on him, and setting him on fire.  No, the arson analysis hadn’t been performed yet, but apparently Castorena had already solved the case.

    Toxicologist Robert Rodriguez, who would actually perform the arson analysis, told me how upset he was because Sonia Cacy had killed two dogs in the process of committing the murder.  No, the arson analysis hadn’t been performed yet, but apparently Rodriguez, just like Castorena, had already solved the case.

    So much for the notion of the fair and unbiased scientist.  The person who would perform the analysis, and the person who would interpret the results of that analysis, were both sure that Cacy had murdered her uncle by setting him on fire.  Before the analysis had even been performed.  Well, that’s how things worked at the Bexar County Forensic Science Center (BCFSC).

    Anyway, the arson analysis was performed.  The analytical results were an unambiguous “none detected.”  No gasoline.  Fine.  And that’s where this whole thing should have ended.  But it didn’t.  Assistant Chief Toxicologist Joe Castorena pronounced that the analysis had indeed determined the presence of gasoline.  He even wrote up a report saying so.  And that’s where this whole thing really begins.


    After Castorena made his pronouncement that gasoline had been found in the sample, I took a look at the analytical results, the actual gas chromatography/mass spectrometry (GC/MS) printouts.  I knew the idea behind the BCFSC’s method of identifying an accelerant.  Both toxicologist Rodriguez and Jeff Todd (the lab chemist who’d been performing arson analyses before the task was handed off to Rodriguez) had explained to me that the identification of an accelerant was based on “pattern recognition.”

    Pattern recognition.  Looking for gasoline in a crime-scene sample?  Analyze a sample of some genuine gasoline, and then analyze the crime-scene sample.  Then compare the results of the analyses.  Gasoline is a mixture of many compounds.  The genuine gasoline sample will show a bunch of peaks, each corresponding to one of the compounds in gasoline.  There will be tall peaks and short peaks, peaks close together and peaks far apart.  Compare those peaks with the crime-scene sample.  Do you get the same pattern of peaks in the crime-scene sample as you did with the sample of genuine gasoline?  That’s pattern recognition.

    So I figured I’d do some pattern recognition of my own.  BCFSC case number1578-91.  I looked at the GC/MS chromatograms.  I compared the chromatogram of the genuine sample of gasoline with the chromatogram of the remnants of clothing taken from Bill Richardson’s body at autopsy.  The clothing sample results looked nothing like the genuine gasoline sample results.  No match.  No gasoline.

February, 1993.

    Sonia Cacy stood trial for the murder of her uncle.  Joe Castorena got up on the witness stand and, under oath, testified that the arson analysis had detected gasoline.  It wasn’t true.  But the jury obviously believed him.  Cacy was found guilty of murder and sentenced to 55 years in prison.

April, 1996.

    Sonia Cacy got a retrial on punishment.  Again, Castorena testified under oath about gasoline being detected in the arson analysis, even though it hadn’t been.  The new jury came back with a new punishment.  55 years wasn’t enough.  Cacy would get the max: 99 years in prison.

Time Goes By.

    While Sonia Cacy was losing years of her life in prison due to Castorena’s fabricated test results, something else started happening.  Dr. Gerald Hurst, an arson and fire-science expert, who had testified at Cacy’s punishment trial on her behalf, began beating the drum: Castorena’s determination of gasoline was a sham, a fiction.  Hurst beat the drum loudly.  One after another, arson experts started looking at the results of the BCFSC analysis.  All said that the analysis had not detected gasoline.  Not one agreed with Castorena.

    Castorena’s false report and his false courtroom testimony started going national.  In July of 1997, the Wall Street Journal ran a story on the Sonia Cacy debacle.  In December of that same year, ABC World News Tonight ran a segment about it.  There was this close-up of a nervous-looking Castorena in his white lab coat.  He says: “I’m one hundred percent certain, uh, you know, of my interpretation of those results.”  And then, after mentioning how the BCFSC had not passed the tests required to be accredited by the American Society of Crime Lab Directors, reporter James Walker went on: “ABC News showed Castorena’s scientific analysis to seven of the nation’s top forensic chemists.  All said he was wrong: there was no evidence of gasoline.”

    And then, in late 1998, Dateline NBC began investigating the Sonia Cacy affair.  During their investigation, Sonia Cacy was magically granted parole.  She was free.  But she’d lost six years of her life thanks to Castorena’s testimony.

    So Cacy was free.  But free on parole.  Cacy was still a murderer in the eyes of the law.

November, 2010.

    Nineteen years.  Nineteen years since Castorena had issued his false report.  Nineteen years since Sonia Cacy’s life had been ruined by Castorena’s false report.  And then something happened: Castorena wrote a letter to Mr. Gary A. Udashen, dated November 30, 2010.  An eleven-page letter, plus attachments.

    Udashen was an attorney working on behalf of Sonia Cacy.  Castorena wrote to Udashen to explain his (Castorena’s) nineteen year insistence that gasoline had been found in a sample where no gasoline had been found.  Castorena’s challenge was to refute the long list of experts who had all said that he was wrong.


    Gasoline.  As I mentioned earlier, gasoline is not a pure substance; it’s a mixture of many compounds (aka “components”).  The GC/MS separates all these compounds and yields a pattern of peaks, each peak corresponding to one of the components.  The result, the “chromatogram,” (aka “chromatograph”) is something like a fingerprint in that you can determine the presence of gasoline in an unknown sample if it shows the same pattern of peaks.

    But the peaks that Castorena had identified as components of gasoline in the sample of clothing did not match the peaks in the genuine gasoline sample.  No “pattern recognition.”  In his letter, Castorena dismisses pattern recognition.  All the experts who had said he was wrong were relying on pattern recognition.  Forget about pattern recognition.  Castorena had his own way of identifying gasoline.  He based his determination of gasoline on the presence of “the C3 alkylbenzene four-peak group.”

    So, what is the C3 alkylbenzene four-peak group?  It’s a group of four peaks found in the GC/MS chromatogram of gasoline.  Castorena claimed to have found the C3 alkylbenzene four-peak group in the clothing sample.  That was good enough for him.  Of the many compounds typically present in gasoline, the C3 alkylbenzene four-peak group was, for Castorena, sufficient for an ID.  According to Castorena, the C3 alkylbenzene four-peak group is “the most important criteria [sic]” for the identification of gasoline.

    Please, as you read this, remember “the C3 alkylbenzene four-peak group.”  It will come up again shortly.


    That doesn’t wrap up Castorena’s November 30, 2010, eleven-page letter to Udashen.  Absolutely not.  On page eight Castorena writes the most incredible thing: “I have indisputable evidence that the Bill Richardson sample was contaminated with xylenes at the morgue.”

    Hold on.  Let’s think about that for a second.  A sample awaiting analysis for gasoline is contaminated with “xylenes.”  Question: what is (or are) “xylenes”?  There are three xylenes: ortho-Xylene, meta-Xylene, and para-Xylene, three different compounds.  (That explains the plural.)  So when Castorena wrote that the Richardson sample had been contaminated with xylenes, he apparently meant all three compounds.

    Here’s the problem: the xylenes are major components of gasoline.  Take a sample that contains no gasoline, contaminate that sample with xylenes, and bingo, you have a sample that looks like it contains gasoline.

    But how can that be if the xylenes are only three compounds and gasoline contains many compounds?  Here’s how: What’s sold as “Xylenes” is a complex mixture of many compounds.  Here’s a list:

1. A large assortment of non-aromatic hydrocarbons
2. Benzene
3. Toluene
4. Ethylbenzene
5. para-Xylene
6. meta-Xylene
7. ortho-Xylene
8. C3 alkybenezenes

All of those compounds found in “Xylenes” are also found in gasoline.  So contaminating a sample that does not contain gasoline with “Xylenes” will make that sample look as if it contains gasoline.  And here’s the biggie: “Xylenes” contains the C3 alkylbenzenes.  Yes, that group of compounds that gives rise to the “C3 alkylbenzene four-peak group” when analyzed by GC/MS.

    Remember the “C3 alkylbenzene four-peak group”?  According to Castorena, “the most important criteria [sic]” for the identification of gasoline?  Remember how all the arson experts were wrong because they were focusing on pattern recognition, while only Castorena was right because he was focusing on the C3 alkylbenzene four-peak group that he claimed to have found?  Castorena even went on to say in his letter that all those arson experts: “… were not privy to all the facts surrounding the analysis of the Bill Richardson’s [sic] sample, namely, contamination in the morgue.”  Ah, yes.  Castorena had secret knowledge.

    Well, duh.  If Castorena really did identify the C3 alkylbenzene four-peak group, it was due to the “Xylenes” contamination that he had “indisputable evidence” of.  Castorena can’t have it both ways.  He can’t say that the C3 alkylbenzene four-peak group is due to gasoline, and then turn around and say that the sample was contaminated with “Xylenes,” which contains C3 alkylbenzenes.  (Well, he can say it.  In fact he has.  But it makes absolutely no sense.  Maybe Castorena is, or was back in 2010 when he wrote that letter, unaware of the chemical composition of what’s sold as “Xylenes.”)  Whatever the case, his alleged finding of gasoline, based on his alleged identification of the C3 alkylbenzene four-peak group, is totally destroyed by his claim of xylenes contamination.

    And wait a second.  Castorena said: “I have indisputable evidence that the Bill Richardson sample was contaminated with xylenes at the morgue.”  The sample was contaminated?  How?  There are really only two possibilities: intentional contamination or non-intentional contamination.  Somebody intentionally contaminated the Bill Richardson clothing sample, a sample that was due to be tested for gasoline, with a chemical mixture that is very much like gasoline.  Or, the clothing sample that was due to be tested for gasoline was somehow contaminated, via some kind of accident or mistake in the morgue, with a chemical mixture that is very much like gasoline.  Some kind of amazing coincidence.

    But nowhere in his letter does Castorena mention the nature of his “indisputable evidence.”  He claims that he has the indisputable evidence, but he doesn’t say what it is.  Nor does he say a word about how the contamination actually occurred.

    And a few huge questions: Did Castorena know about this alleged contamination back in 1991when he’d claimed that he found gasoline in the clothing sample?  If yes, why did he wait nineteen years to open up about it?  If no, when and how did he learn of it?

January, 2014.

    An exoneration hearing (aka “writ hearing”) for Sonia Cacy was approaching - a hearing to determine whether or not there was sufficient reason to have that 1993 finding of “guilty” changed to “not guilty.”  The District Attorney’s office wanted to get an arson expert on its side, someone who would back up Castorena’s false claim of the presence of gasoline.

    Hey, if you’re trying to get an expert to refute the unanimous opinion of a whole bunch of other experts, you’d better get somebody good.  They contacted Dr. Elizabeth Buc.  Ph.D. in Materials Science and Engineering.  M.S. and B.S. in Chemistry.  Certified Fire Investigator.  Definitely somebody good.  They sent her the GC/MS chromatograms from the BCFSC arson analysis and asked her for her opinion on the determination of gasoline in the Bill Richardson clothing samples.

    In a January 23, 2014 letter to Mr. Donald McCarthy, Esq., Assistant District Attorney in Alpine, Texas, Dr. Buc wrote: “It is my opinion, based on this review, on my education in chemistry and materials science and engineering and on my experience in fire investigation and fire science, that the chromatographs are negative for gasoline.”

    Ouch.  That didn’t work out very well.  Another expert saying that Castorena was wrong, that there was no gasoline detected in the sample of clothing taken from Bill Richardson’s body at autopsy.  Only this time, the expert had been hired by the folks trying to show that Castorena was right.

June, 2014.

    Sonia Cacy got her writ hearing.  Guilty of murder.  Or not guilty of murder.  That’s what was hanging in the balance.

    Joe Castorena was subpoenaed to testify.  Quickly, Castorena’s story about the arson analysis got weirder than it already was.  Remember the contamination with xylenes in the morgue?  Well now Castorena was saying that there was additional contamination.  This time the sample had also been contaminated with toluene in the toxicology lab.  (Toluene is a component of gasoline.)

    So let me get this straight: A sample due for an analysis to determine the presence of gasoline was contaminated with components of gasoline (xylenes) in the morgue, and then the sample was walked down the hall in a sealed container to the toxicology lab, where it was contaminated a second time, this time with another component of gasoline (toluene).  If it weren’t for what happened to Sonia Cacy, this would actually be funny.

    Gary Udashen (Cacy’s attorney) asked Castorena when the contamination was discovered.  Castorena’s reply: “Oh, we knew about this contamination - I couldn’t tell you exactly when, but it was definitely well before the analysis of the Bill Richardson sample.”  Castorena then narrowed down that “well before” to “several months.”

    Udashen caught that “we” in “we knew” and asked: “Who’s ‘we’?  You and who else?”

    Guess who.  Correct.  Toxicologist Robert Rodriguez, the person who’d actually performed the analysis on the sample.  Like Castorena, Rodriguez also possessed the secret knowledge.  The only two people in the entire BCFSC who knew about those “several months” of contamination were Castorena and Rodriguez.  Did they tell anyone else?  Maybe the medical examiners?  Or the other people working in the toxicology lab?  Nope.

    Okay.  During those “several months” of contamination, were any other arson samples contaminated?  Castorena: “I would say at least ten or more.”  Did Castorena ever note on a report, or mention when testifying at trial, or mention to an attorney, a judge, or anyone else about the contamination?  Nope.  When asked if not telling anyone about the contamination “was a significant error in judgment,” Castorena replied: “No, it wasn’t.  Nobody asked me, you know, about that possible problem.”  Pushed further on this, Castorena went on: “I said nobody asked me, so that’s why I never told anyone about this situation.”  (This “situation” being contamination of arson samples with components of gasoline.)

    Udashen asked Castorena if he thought it was appropriate for a forensic scientist to not reveal problems with an analysis and testing methods.  Castorena: “Well - I’ve been trained to - you know, in testifying, to answer the question.  And if I’m not asked the questions, then, you know, I don’t - I don’t answer that.”

    Udashen asked Castorena if, had he not known about the contamination, would he have agreed with all the other experts that no gasoline had been detected.  Castorena: “I would have concluded an undetected, yes.”  He would have agreed.  No gasoline.

    Udashen asked about those ten-plus cases that had been contaminated during those “several months” of xylenes/toluene contamination in the morgue and toxicology lab.  Had Castorena yet told anyone about the contamination?  Castorena: “No.”

    Udashen asked Castorena if he was familiar with the report of Dr. Elizabeth Buc, the arson expert contacted by the District Attorney in the hopes of finding that one expert who might agree with Castorena’s false determination of gasoline.  Buc, the arson expert who’d said back in January, “…the chromatographs are negative for gasoline.”  Castorena claimed that he couldn’t remember if he’d actually seen her report, but he thought that she had not agreed with his finding of gasoline.  Question: Why didn’t she find gasoline?  According to Castorena: same problem as everybody else.  She didn’t know about the contamination.  She didn’t have the secret knowledge.

    Castorena testified that he did not share his secret knowledge with Dr. Buc, the expert hired by the DA to support Castorena’s own findings.   (But Castorena had revealed his secret knowledge in a letter sent years earlier to Sonia Cacy’s lawyer.)  And Castorena had already testified that without the secret knowledge, he himself would have made a call of “none detected.”  Dr. Buc was doomed to find no gasoline.  She would just be another expert coming to the same wrong conclusion.

    And what about the whole contamination problem?  Whatever happened to it?  Did it just go away by itself?  Yeah, there was this period of “several months” where arson samples were being mysteriously contaminated with xylenes in the morgue and toluene in the toxicology lab, but then one day it just stopped.  That’s it?  Magic?

    Udashen asked: “And when was the contamination fixed?”  Castorena: “Again, I certainly can’t remember when that happened, but at one point well after - well after the Richardson sample was analyzed, we - we were able to solve the problem.”  There’s that “we” again.  But since nobody else on the planet knew about the contamination problem until 2010, Castorena and Rodriguez must have figured out the source of the problem and fixed it all by themselves.  After hours?  Screwdrivers and flashlights and duct tape?

June, 2016.

    The Honorable Bert Richardson ruled that Sonia Cacy was not guilty.  More than twenty-three years after having been found guilty, Cacy was no longer guilty in the eyes of the law.  Twenty-three years.  All ruined because of Castorena’s false report and false testimony.

Sonia Cacy

In Closing.

    1) Those ten or more arson cases Castorena mentioned in his testimony, the ones that had been contaminated in the morgue and toxicology lab.  Shouldn’t someone be reexamining them?

    2) That whole contamination thing.  What’s that all about?  Castorena states that there was this ongoing (“several months”) problem with xylenes (in the morgue) and toluene (in the toxicology lab) contaminating arson samples.  How?  At no point does Castorena explain the nature of these contaminations, the mechanisms behind them.  But he testifies that there was some ongoing problem in the BCFSC that he and Rodriguez had finally, after many months, solved.  What did they do?

    3) Mysterious contamination in the morgue?  And mysterious contamination in the toxicology lab?  Critical details that Castorena and Rodriguez, forensic scientists, never shared with anyone until Castorena’s 2010 letter?  I don’t buy it for a second.  If Bill Richardson’s clothing sample had truly been contaminated with xylenes and toluene, a much simpler and much more reasonable explanation is that someone at the BCFSC, someone who had access to the Bill Richardson clothing sample, deliberately contaminated it with xylenes and toluene in an attempt to make it appear as if the sample were positive for gasoline.

    4) Did Toxicologist Robert Rodriguez really know about the alleged contamination in the lab?  And like Castorena, he never said anything to anyone about it?  Hm.  I don’t think so.  I knew Rodriguez.  We were friends.  We went to lunch together.  We told each other jokes.  I knew his wife.  I went to his house.  He knew my wife.  He came to our house.  I can’t see Rodriguez keeping his mouth shut about something so important.


    So anyway, did someone in the BCFSC deliberately contaminate the sample?  I don’t know.  That’s all I can say about it: I don’t know.  But I did spend four years at the BCFSC, the Crime Lab from Hell, and nothing would surprise me.

    One last thing.  At Sonia Cacy’s writ hearing, Gary Udashen asked Castorena: “Prior to hearing from me, had you kind of forgotten about Sonia Cacy?”  Castorena’s reply: “Hadn’t given it much thought, no.”

In memory of Dr. Gerald Hurst, 1937 - 2015.
Larry Ytuarte, Ph.D., was born and raised in New York City.  He earned a bachelor’s degree in chemistry and philosophy from York College of the City University of New York and a Ph.D. in chemistry from Brown University.  He worked for 4 years as a forensic chemist at the Bexar County Forensic Science Center in San Antonio, Texas.  He currently lives in Las Cruces, New Mexico, with his wife, Louise.

Junk Science

Truth in Justice