Sybers v. State

841 So.2d 532 (2003) William SYBERS, Appellant, v. STATE of Florida, Appellee. Nos. 1D01-1609, 1D02-129. District Court of Appeal of Florida, First District. February 28, 2003. *534 Nathan Z. Dershowitz and Amy Adelson of Dershowitz, Eiger & Adelson, P.C., New York, NY; Alan M. Dershowitz, Cambridge, MA; Lorence Jon Bielby of Greenberg Traurig, P.A., Tallahassee; and Elliot H. Scherker of Greenberg Traurig, P.A., Miami, for Appellant. Charlie Crist, Attorney General; Robert R. Wheeler, Assistant Attorney General, Tallahassee; and Harry L. Shorstein, Special Assistant Attorney General, Jacksonville, for Appellee. WEBSTER, J. In these consolidated appeals, appellant seeks review of his conviction, following a jury trial, for first-degree murder and the denial of his motion for postconviction relief. On direct appeal, appellant contends that the trial court committed reversible error when it admitted, following a Frye hearing, expert testimony based on tests purportedly establishing the presence of succinylmonocholine in the victim's embalmed tissue nine years after the victim's death; and when it allowed the state to establish a motive for murder through the use of inadmissible testimony. We agree. Accordingly, we reverse, and remand for a new trial. This disposition moots appellant's postconviction claims. I. A. At all relevant times, appellant was the medical examiner for the Fourteenth Judicial Circuit of Florida. On the morning of May 30, 1991, appellant's wife was found dead in their home. Appellant had asked one of his employees to check on his wife because, according to appellant, his wife had been complaining of chest and shoulder pains that morning, and nobody was answering the telephone at home. Appellant's wife was found in her bed, with the covers over her and a heating pad on her leg. She had no pulse. There were no signs of a struggle, but there was a gauze bandage on her arm. When appellant was told of his wife's death, he said that no autopsy should be performed because his wife "wouldn't want that." The next morning, appellant's medical partner persuaded him that an autopsy would be appropriate. However, by that time, the body had already been embalmed by the funeral home. Appellant had not sought to expedite the embalming. *535 An associate medical examiner from the First Judicial Circuit performed the autopsy on June 1, 1991. He noted two injection marks on the right arm. (Appellant had told at least two people that he had tried unsuccessfully to draw blood from his wife early on the morning of her death, because she had not been feeling well.) The only drug discovered by a toxicology screen was a sleeping pill. The medical examiner was unable to find any signs of disease or injury that might explain the death. On October 16, 1991, he issued a death certificate listing "sudden unexpected death due to undetermined natural causes." Two weeks later, he issued another death certificate, this time listing "sudden unexpected death due to undetermined causes." On April 20, 2000, he issued a final death certificate listing the death as a homicide due to succinylcholine poisoning. The final death certificate was based on reports received from National Medical Services and the Federal Bureau of Investigation. Appellant was indicted for first-degree murder on February 18, 1997. The indictment charged that appellant had murdered his wife "by injecting her with an unknown substance." On April 9, 1997, in response to a story in the local newspaper quoting the prosecutor, appellant filed a motion in limine seeking to exclude "any evidence concerning or reference to potassium or alleged potassium poisoning, and any novel scientific evidence." In that motion, appellant maintained that, in Florida, "[t]he proponent of scientific evidence must satisfy the test announced in Frye v. United States, 293 F. 1013 (D.C.Cir.1923), ... [as] further refined by the Florida Supreme Court ... in Brim v. State, 663 So.2d 629 (Fla.1995)[sic]"; that "[t]he burden is on the proponent of the evidence to prove the general acceptance of both the underlying scientific principle and the testing procedures used"; and that he was "entitled to a pretrial hearing to determine the admissibility of any proposed novel scientific evidence." On November 19, 1998, following a Frye hearing, the trial court held that opinion testimony offered by the state from Dr. Frederic Rieders, of National Medical Services (NMS), a private laboratory in Pennsylvania, was inadmissible because it was "based upon methods and processes that [we]re not generally accepted... in the scientific community and therefore d[id] not meet the Frye test." The state sought certiorari review in this court. On October 20, 1999, we issued a short opinion, in which we agreed "that the state failed to establish the necessary scientific acceptance as to some of the principles and methods on which [Dr. Rieders] relied" and, accordingly, denied the petition for a writ of certiorari. State v. Sybers, 743 So.2d 619 (Fla. 1st DCA 1999). B. On January 11, 2001, appellant filed another motion in limine, this time seeking to exclude "any evidence concerning or reference to succinyl choline poisoning or succinyl monocholine poisoning, and any novel scientific evidence." A pre-trial Frye hearing was held on February 14 and 15, 2001. At that hearing, Dr. Kevin Ballard testified for the state. He said that he was the director of research and development at NMS. Although he had never practiced medicine, he had a medical degree as well as a Ph.D. in pharmacology. He had been declared in other courts to be an expert in tandem mass spectrometry (MS/MS). He had no formal training in testing embalmed tissue, but had been involved in the testing of embalmed tissue specimens since 1998. He initially became involved in appellant's case in December 1999, at the request of Dr. Rieders. According to Dr. Ballard, he had developed a "bench procedure" for quaternary *536 ammonium compounds in biological specimens which he used to test embalmed tissue from appellant's deceased wife for the presence of succinylcholine, a neuromuscular blocking agent commonly used during surgery, which is lethal in sufficient doses. The testing procedure was a multi-step process. The first step was to homogenize the tissue sample in water to render it in a form that could be worked with. The second step was to begin the extraction of the analyte of interest from the homogenized sample through use of the Bligh-Dyer method, a liquid-liquid extraction procedure dating back to the 1930's, which removed lipids and left only water-soluble components. The final step was a solid phase extraction procedure which used an ion pairing reagent, heptafluorobutyric acid, to isolate peptides from biological samples, yielding a final liquid matrix suitable for liquid chromatography. Dr. Ballard testified that his bench procedure was unique insofar as it combined the Bligh-Dyer method and the ion-pairing solid phase approach during the extraction process. Once the final liquid matrix was extracted, the matrix was injected into a liquid chromatograph-tandem mass spectrometer (LC-MS/MS). The chromatograph separated the compound, and the mass spectrometer identified it. On February 17, 2000, Dr. Ballard reported that succinylmonocholine (a metabolite formed by the decay of succinylcholine) was detected in the victim's kidney, brain, and lung samples using LC-MS/MS on a tandem quadrupole mass spectrometer. On March 22, 2000, he prepared another report which detected succinylmonocholine in the victim's spleen, kidney, brain, lung, and fat samples again using LC-MS/MS on a tandem quadrupole mass spectrometer. On November 7, 2000, he issued a supplemental report reflecting an additional analysis of the tissue samples using LC-MS/MS on a tandem quadrupole-time-of-flight instrument. This procedure showed the presence of succinylmonocholine in the kidney and brain specimens. In June 2000, he made a "poster" presentation to the American Society for Mass Spectrometry entitled "Analysis of Quaternary Ammonium Neuromuscular Blocking Agents in Forensic Tissue and Fluid Specimens by LC-MS/MS." It involved taking questions from any peers who came by to examine the presentation. The presentation did not detail the bench procedure used, but did identify the concepts for extracting and identifying the compounds. Dr. Ballard testified, further, that he did not perform method validation on the victim's specimen samples, as that term was traditionally used, because it was impossible to do a true validation study on unique specimens. Method validation was intended for routine testing. Instead, he validated his methodology by using the "standard addition" method, which was generally accepted in the scientific community. This involved taking a portion of a specimen and "spiking" it with the analyte of interest to show the capability of pulling that analyte out of the matrix. This was called the positive control. The standard addition method also involved the use of a negative control, i.e., a "blank" sample known not to have the analyte in it, to ensure against false positives. One cause of false positives was "carryover," in which traces of analytes injected into or extracted from high level samples would remain in the equipment and contaminate subsequent samples. The solution to this problem was to run a number of "blank" samples between runs to make sure that the system had been properly purged. They never experienced any carryover with succinylcholine or succinylmonocholine. Another source of contamination was "interference," in which some unknown compound could potentially *537 mimic or mask the analyte being targeted. This was controlled by use of the standard addition method. Interference would not cause a false positive unless there was another precursor, other than succinylcholine, that could break down and give rise to succinylmonocholine. Theoretically, there was some possibility that embalming fluid could cause a false positive, but this was so small as to be essentially ignorable. The state also presented the testimony of Marc LeBeau, unit chief supervisory chemist in the toxicology department of the Federal Bureau of Investigation (FBI) laboratory. Mr. LeBeau testified that he had experience using tandem mass spectrometry with embalmed tissues, and in identifying succinylcholine. In December 1999, he received specimens related to this case from NMS. He also received a copy of NMS's bench procedure which was conditioned on the FBI signing a nondisclosure agreement. He performed five tests on the specimens: LC-MS full scan; LC-MS/MS using selective reaction monitoring (SRM); LC-MS/MS full scan; LC-MS/MS/MS using SRM; and LC-MS/MS/MS full scan. The LC-MS technique had been around for a few decades, but its application to forensic science was fairly new. That technique involved taking a liquid and immediately putting it into a gaseous state. The MS/MS and MS/MS/MS techniques provided unique structural information about the analytes being looked for, and were much more selective and sensitive for most applications than the MS technique. Moreover, the LC-MS/MS/MS method was a more effective test for succinylcholine or succinylmonocholine than the older gas chromatography-mass spectrometry method traditionally used. LeBeau testified that he performed two separate procedures using the specimens he received. The first was an extraction procedure involving a combination of filtration, liquid extraction with an ion pairing reagent, and solid phase extraction. He then analyzed the extract using an instrumental procedure which involved a combination of LC-MS/MS and LC-MS/MS/MS. His testing indicated a positive finding for succinylmonocholine in the victim's kidney, which finding was reproduced by two other toxicologists working under his supervision. This was a qualitative rather than a quantitative analysis, i.e., it answered the question of whether a substance was present, rather than how much was present. To validate the testing protocol, he used the standard addition method, which was the FBI laboratory's standard approach when dealing with unique specimens such as embalmed tissue. He tested a number of embalmed tissue samples that he knew did not contain succinylmonocholine to make sure that there were no outside contaminants and that succinylmonocholine was not a naturally occurring product in embalmed tissue. He never produced a positive result from a specimen known not to contain the drug. He tested a number of embalmed tissue samples that were spiked with succinylmonocholine to establish that the analyte could be extracted from the tissue. He performed it on different embalmed tissue samples from the victim because a kidney would not necessarily react the same as a liver. Every single specimen was different from the next, and some might not test positive because the matrix was unique and embalmed tissues were difficult to work with. In addition, succinylmonocholine was a difficult chemical to work with because it was unstable and disappeared if the temperature was too hot or the pH level was too extreme. Because of this instability, it could be found in one sample and not in another. To establish that the extraction procedure did not introduce anything that would be misidentified as succinylmonocholine, he performed an interference *538 study by taking plain water and carrying it through the procedure to make sure it was working properly. He looked at common putrefaction products and discerned that none would interfere with the results. He looked at other chemicals with similar weights to succinylmonocholine and determined that they would not give the same reading on the mass spectrometer. He also performed a carryover study to ensure that the instrument was appropriately flushed out after each use. According to Mr. LeBeau, there might be a number of reasons why Dr. Ballard identified succinylmonocholine in more samples. His detection limit might have been higher than Ballard's. The drug might have been present, but not detected because of the higher detection limit. Different detection limits resulted from the use of different instruments. Another reason for the different findings might have been the instability of succinylmonocholine. The homogenization process, i.e., putting the tissue into a blender, created heat which might have caused the succinylmonocholine to break down. He knew that he did not produce a false-positive finding because of the validation procedure. Given the complexity of mass spectrometry, it was essentially impossible to detect anything other than succinylmonocholine. LeBeau testified, further, that the uniqueness of the NMS bench procedure was not in the extraction. Liquid extraction and solid phase extraction were not unique. He used those techniques on a daily basis. Ion pairing was in all of the literature dealing with quaternary ammonium compounds, and was a routine technique to isolate the compounds out of an aqueous matrix. The NMS extraction procedure was an older procedure called the Bligh-Dyer procedure. NMS considered unique the combination of the extraction and instrumental procedures. It was one of the only laboratories capable of doing both. Its instrumental procedure using LC-MS/MS was also unique. LeBeau said that he performed the same extraction process as NMS, and validated it. However, he did a completely different instrumental analysis, which had been peer reviewed by the Society of Forensic Toxicology and within the FBI lab by other FBI scientists. All of the samples were prepared in essentially the same manner. He did not "tweak" the specimens. He "tweaked" the instrumentation by making minor modifications so that it had a better chance of determining whether something was there. Although there were some "deviations," "none of the deviations had any scientifically valid detriment." The only specimen he can remember "tweaking" was the liver specimen, by adding more solvent, because it was badly decomposed. Appellant presented the testimony of Janine Arvizu, a chemist and a laboratory quality consultant; Dr. Ashraf Mozayani, the chief toxicologist and lab director of the Houston, Texas, medical examiner's office; and Dr. Graham Jones, the director of the medical examiner's toxicology laboratory in Edmonton, Alberta. Drs. Mozayani and Jones testified that LC-MS/MS was the best available method for detecting the presence of succinylmonocholine in embalmed tissue and that the individual steps in Dr. Ballard's bench procedure— tissue homogenization, liquid-liquid extraction, solid phase extraction with ion pairing, and LC-MS/MS—were generally accepted within the scientific community. However, they claimed that Ballard's novel application of these procedures had not been validated, and that method validation was a prerequisite to acceptance in the scientific community. During method validation, the analyst could not change any *539 procedural steps, as was done by NMS, because any change interfered with the validation process. Dr. Mozayani testified that Dr. Ballard's reason for bypassing validation—that the specimens were unique—was not sound, because Dr. Ballard could have used other embalmed tissue specimens to validate his methodology. She and Dr. Jones also criticized the use of the standard addition method because it lacked adequate controls. Both witnesses also expressed concerns about the failure of Dr. Ballard and Mr. LeBeau to publish their work, or subject it to peer review. They claimed that the "poster" presentation by Dr. Ballard at the American Society for Mass Spectrometry did not constitute peer review because it did not present Ballard's bench procedure, did not allow for reproduction of his methodology, and was not subjected to critical analysis. Another concern involved Dr. Ballard's failure to have his methodology "reproduced" by an independent laboratory. Drs. Mozayani and Jones testified that Dr. Ballard's methodology could not be "reproduced" by the FBI because the FBI used a different methodology and was not able to reproduce Dr. Ballard's results, detecting succinylmonocholine in only one specimen. Appellant's experts testified regarding other alleged deficiencies which precluded a finding of general acceptance in the scientific community. According to them, NMS made no attempt to determine a known error rate, and the FBI made only an ineffective attempt. Moreover, neither NMS nor the FBI documented "limits of detection" (the lowest concentration that can be detected in a single measurement of a single sample), which was critical because the FBI detected only a trace amount of succinylmonocholine in only one specimen, and there was no literature on whether succinylmonocholine was naturally found in trace amounts in the body. Dr. Jones testified that he did not consider the FBI's findings to be particularly compelling since the level of succinylmonocholine was so close to the limit of detection. In addition, Drs. Mozayani and Jones testified that there was no scientific literature or study demonstrating either that succinylmonocholine was not present in embalmed tissue or that succinic acid and choline, both of which were endogenous to the human body, could not recombine to form succinylmonocholine. They maintained that this was significant in this case because the state-of-the-art instruments used by NMS and the FBI were extremely sensitive, and capable of picking up compounds in very minute concentrations. Thus, if there were traces of endogenous succinylmonocholine in embalmed tissue, it would be detected by their sensitive instruments. The witnesses also testified that there must be protection against contamination at each stage of the process. According to Ms. Arvizu, NMS had no quality assurance program in place to ensure sample integrity or to prevent contamination. One key aspect of quality assurance involved including negative control samples to ensure against false positives. Ms. Arvizu observed that Dr. Ballard did not always include negative controls in his testing. In at least one instance, she noted that there was a positive reading for succinylmonocholine when the lab ran a blank sample. She also noted that because of poor record keeping, there was no way to identify the samples that were tested during particular analytical runs. The lab had no system in place to restrict access to the forensic sample preparation area. Dr. Rieders reported that when he removed a sample from the freezer storage area, he did not record it on the custody control documentation. The lack of quality controls was significant because NMS was involved in testing tissue *540 samples in another criminal case that were known to contain succinylcholine. Ms. Arvizu also testified that the FBI had a general quality assurance protocol, but did not follow it in this case. Dr. Jones was particularly concerned that the FBI testing reflected unusual carryover. He noted that even after the instrument was purged and blank samples were repeatedly tested, carryover was still detected, a problem not adequately addressed by the FBI. Following the hearing, the trial court denied the motion in limine. It concluded that Dr. Ballard's methodology satisfied the Frye standard because the individual steps in the NMS and FBI procedures were generally accepted in the scientific community and other courts had allowed prosecutors to introduce expert evidence concerning the detection of succinylcholine in embalmed tissue. As to the lack of method validation and quality assurance controls and the possibility of contamination and carryover, the trial court concluded that, pursuant to Frye, such matters went to the weight of the evidence, rather than to its admissibility. C. During the trial, appellant again moved to exclude the state's expert opinion testimony regarding the presence of succinylmonocholine in the victim's embalmed tissues. Ms. Arvizu and Mr. Chip Walls, a forensic toxicologist, both testified that, based upon their observation of NMS procedures, NMS's forensic processing area did not have in place even minimal elements of a quality assurance program. As a result, there was a significant possibility of contamination of the tissue samples. Arvizu also expressed grave concerns regarding NMS's chain of custody documentation for the samples, which she said raised substantial questions as to the identity of the samples. The trial court again denied the motion, stating that any problems regarding contamination or chain of custody went to the weight, rather than the admissibility, of the evidence. Dr. Ballard testified that he used his bench procedure and LC-MS/MS to conduct three tests on embalmed tissue samples taken from appellant's wife. The first detected succinylmonocholine in the kidney, brain, and lung samples; the second detected succinylmonocholine in those organs and the spleen and fatty tissue; and the third (using newer and more sensitive equipment) detected succinylmonocholine in only the brain and kidney. More than trace amounts were detected. According to Ballard, the presence of succinylmonocholine established the presence of its parent drug, succinylcholine. Mr. LeBeau testified that he conducted tests on the specimens received from NMS using Ballard's bench procedure, with some modifications, and the FBI's more selective LC-MS/MS/MS instrument, and found succinylmonocholine only in the kidney sample. Both he and Ballard theorized that succinylmonocholine could be detected in the tissues of appellant's wife, despite the instability of succinylmonocholine and the passage of time, because embalming acidified the tissue, a condition in which succinylmonocholine was more stable. However, Ballard acknowledged that he did not know the chemical makeup of the embalming fluid. LeBeau testified that his results differed from those of NMS because of different methodology and instrumentation, because the specimens had been thawed, or because of differences in the homogenization procedures. Drs. Mozayani and Jones, and Mr. Walls, testified that succinylcholine rapidly degraded and was generally out of the body in less than ten minutes while succinylmonocholine *541 degraded eight times more slowly because of its different chemical structure. Mozayani and Walls testified, further, that embalming could vary the process depending on whether the embalming fluid contained sodium hydroxide, which would cause the succinylcholine to degrade more quickly, or an acid, which would stabilize it. They cited a study published in the Journal of Forensic Science involving degradation rates of succinylcholine in canine kidney tissue which established that a body must be immediately placed in minus 70 degrees Celsius to prevent degradation of succinylcholine. In 40 days, 78 percent of succinylcholine was gone at minus 20 degrees Celsius, and 91 percent was gone in refrigerated temperatures. Using LeBeau's detection limits, Mozayani and Walls agreed that any succinylmonocholine would have degraded by the time NMS conducted its tests even if the tissue samples had been stored in a freezer. In their opinion, succinylmonocholine could not be detected within the tissues after two years when the tissues were embalmed 90 minutes after death and immediately preserved in a freezer. Moreover, they agreed that there was no reasonable explanation why succinylmonocholine was found in the brain, spleen, lung and fatty tissue in December 1999, but could not be found by the FBI in the same tissue 61 days later. If the compound degraded so quickly between the NMS and FBI testing, Dr. Jones testified that this was inconsistent with the tests being positive in the first place. D. Also during the trial, the state called Judy Ray Sybers, who married appellant three years after the death of appellant's wife. She testified that she had begun having a sexual relationship with appellant in February 1991, but that she and appellant had not discussed marriage. She acknowledged lying to the police about the affair, but stated that appellant had told her to tell the truth. Another witness, Diane Houser, was permitted to testify, over objection, that about a month before the death of appellant's wife, Judy had told her that she was seeing appellant and that she thought appellant was going to try to get a divorce. Again over objection, David Levin, a divorce attorney, was permitted to testify that under Florida law, marital assets would be divided equally on divorce and that, assuming a net worth of $5-6 million, the wife would receive between $2½-3 million. II. A. "[N]ovel scientific evidence is not admissible in Florida unless it meets the test established in Frye v. United States, 293 F. 1013 (D.C.Cir.1923)," which requires that such evidence "`be sufficiently established to have gained general acceptance in the particular field in which it belongs.'" Flanagan v. State, 625 So.2d 827, 828 (Fla.1993) (quoting Frye, 293 F. at 1014). Accord Hadden v. State, 690 So.2d 573, 577-78 (Fla.1997); Ramirez v. State, 651 So.2d 1164, 1167 (Fla.1995) (Ramirez II). Among other things, "[t]his standard requires a determination ... that the basic underlying principles of scientific evidence have been sufficiently tested and accepted by the relevant scientific community." Brim v. State, 695 So.2d 268, 272 (Fla. 1997). As our supreme court has explained: The underlying theory for this rule is that a courtroom is not a laboratory, and as such is not the place to conduct scientific experiments. If the scientific community considers a procedure or process unreliable for its own purposes, then the *542 procedure must be considered less reliable for courtroom use. Stokes v. State, 548 So.2d 188, 193-94 (Fla. 1989). In other words, the Frye test is "designed to ensure that the jury will not be misled by experimental scientific methods which may ultimately prove to be unsound." Flanagan, 625 So.2d at 828 (citing Stokes). When applying the Frye test, "the burden is on the proponent of the evidence to prove [by the greater weight of the evidence] the general acceptance of both the underlying scientific principle and the testing procedures used to apply that principle to the facts of the case at hand." Ramirez II, 651 So.2d at 1168. This requires more than "[a] bald assertion by the expert that his deduction is premised upon well-recognized scientific principles," particularly "if the witness's application of these principles is untested and lacks indicia of acceptability," or "if the expert has a personal stake in the new theory or is prone to an institutional bias." Ramirez v. State, 810 So.2d 836, 844, 844 n. 13 (Fla. 2001) (Ramirez III). As our supreme court has emphasized, "general scientific recognition requires the testimony of impartial experts or scientists. It is this independent and impartial proof of general scientific acceptability that provides the necessary Frye foundation." Id. at 851. Our standard of review when considering a trial court's ruling on a Frye issue is de novo, rather than abuse of discretion. Murray v. State, 838 So.2d 1073, 1077 (Fla.2002); Ramirez III, 810 So.2d at 844; Brim, 695 So.2d at 274. Moreover, we must address the matter of general acceptance as of the time of appeal, rather than the time of trial. Ramirez III, 810 So.2d at 844-45 (citing Hadden, 690 So.2d at 579). In reaching our decision regarding general acceptance, we "may examine expert testimony, scientific and legal writings, and judicial opinions." Hadden, 690 So.2d at 579 (citing Flanagan, 625 So.2d at 828). "Any doubt as to admissibility under Frye should be resolved in a manner that minimizes the chance of a wrongful conviction." Ramirez III, 810 So.2d at 853. B. The scientific evidence in this case involved two basic underlying principles: (1) the detection of succinylmonocholine establishes the prior presence of its parent compound, succinylcholine; and (2) succinylmonocholine can remain in embalmed tissue nine years after death. The evidence is uncontradicted that succinylcholine is a very unstable compound that breaks down rapidly to produce succinylmonocholine, a less unstable compound that breaks down to form succinic acid and choline, which are naturally present in the human body. Appellant's experts testified that succinylmonocholine is too unstable to be present in embalmed tissue nine years after death. The state's experts hypothesized that the chemicals used in the embalming process preserved succinylmonocholine in the tissues by making it more stable. Appellant's experts acknowledged that embalming fluids containing acids might slow the decomposition of succinylmonocholine, but further testified that embalming fluids containing sodium hydroxide would accelerate the degradation process. There is no evidence in the record establishing the chemical composition of the embalming fluids used in this case. Furthermore, no literature or studies have been cited or found addressing the effect of embalming on the decomposition of succinylmonocholine. In addition, appellant advanced two possible explanations for the presence of succinylmonocholine in the embalmed tissues of appellant's wife other than the injection of succinylcholine. The first was the natural *543 recombination of succinic acid and choline in the body to form succinylmonocholine in trace amounts. While acknowledging that there were no scientific studies confirming or rejecting this hypothesis, appellant's experts testified that it was theoretically possible and needed further study. The state's experts, Dr. Ballard and Mr. LeBeau, dismissed this as very unlikely given the energy required to recombine succinic acid and choline. The second explanation advanced by appellant was contamination. We conclude that the state has failed to carry its burden of establishing by "independent and impartial proof that the scientific principles underlying the testing, i.e., that an unstable compound like succinylmonocholine can be preserved in nine-year-old embalmed tissue and that it could come only from an injection of succinylcholine, are generally accepted in the relevant scientific community. Other than the assertions of Dr. Ballard and Mr. LeBeau, the state offered no evidence to suggest that either scientific principle is generally accepted. As the court noted in Ramirez III, such assertions by experts who developed and performed the testing procedures are not, alone, sufficient. Id., 810 So.2d at 844. Rather, proof of general scientific acceptance "requires the testimony of impartial experts or scientists," id. at 851, "that the ... underlying principles... have been sufficiently tested and accepted by the relevant scientific community." Brim, 695 So.2d at 272. The state argues that we should disregard Ramirez III because it was not decided until after the trial had been concluded. We reject this argument for two reasons. First, the requirement that general acceptance be established by "independent and impartial proof" is nothing more than a common-sense conclusion drawn from the supreme court's decisions which preceded Ramirez III. Accordingly, Ramirez III did not announce a new rule of law in this regard. Second, even if it had, any decision of the supreme court "announcing a new rule of law, or merely applying an established rule of law to a new or different factual situation, must be given retrospective application by the courts of this state in every case pending on direct review or not yet final," unless the supreme court states otherwise. Smith v. State, 598 So.2d 1063, 1066 (Fla. 1992), as limited by Wuornos v. State, 644 So.2d 1000, 1007 n. 4 (Fla.1994). The supreme court has not suggested that Ramirez III should not be given retrospective application. C. The actual testing procedure used to detect succinylmonocholine in this case was developed by Dr. Ballard at NMS. The first half of the procedure was an extraction process which involved (1) using a blender to homogenize the tissue sample in water; (2) using a liquid-liquid extraction procedure called the Bligh-Dyer method to remove unwanted lipids (fats) from the sample; and (3) using an ion-pairing solid phase extraction procedure to yield a final liquid matrix suitable for liquid chromatography. The second half of the procedure was an instrumental analysis which involved using liquid chromatography (LC) to separate the analyte of interest from the liquid matrix and tandem mass spectrometry (MS/MS) to identify it. Follow-up testing was performed at the FBI laboratory by Mr. LeBeau, who testified that he performed a similar extraction procedure to that used by NMS, but performed a different instrumental analysis which involved a combination of LC-MS/MS and LC-MS/MS/MS. While appellant's experts agreed that the individual steps of the extraction procedure and the use of LC-MS/MS and LC-MS/MS/MS *544 were generally accepted in the relevant scientific community, they asserted that the application of the extraction procedure and instrumental analysis to isolate and detect succinylmonocholine in nine-year-old embalmed tissue was not generally accepted in the relevant scientific community. No scientific literature has been cited by the parties, or uncovered by our independent research, which addresses similar testing procedures for succinylmonocholine in embalmed tissues, especially tissues that are many years old. The parties have cited three other appellate cases which addressed the testing of post-mortem tissue for the presence of succinylcholine, and the state points out that, in all three, the presentation of expert testimony regarding tests used to detect the presence of the drug was affirmed on appeal. The cases are Coppolino v. State, 223 So.2d 68 (Fla. 2d DCA 1968), Jones v. State, 716 S.W.2d 142 (Tex.Ct. App.1986), and People v. Davis, 199 Mich. App. 502, 503 N.W.2d 457 (1993), leave to appeal denied, 444 Mich. 966, 518 N.W.2d 475 (1994). We find these cases to be of little help. All involved different testing techniques; all involved different testing equipment; all involved testing for succinylcholine, rather than its metabolite, succinylmonocholine; none involved testing of tissue that had been embalmed for a period remotely approaching nine years; and all involved application of a much more permissive interpretation of the Frye test than that which has been mandated by our supreme court. In addition, the court in Coppolino applied the abuse of discretion standard of review, which is much more permissive than the de novo standard since mandated by our supreme court; and the Jones and Davis cases both recite that expert witnesses other than the developers of the tests testified that the tests were generally accepted within the relevant scientific community. We conclude that the state has failed to carry its burden of establishing by "independent and impartial proof" that the testing procedures used are generally accepted in the relevant scientific community. The only testimony offered by the state to establish the general scientific acceptance of the testing procedures came from Dr. Ballard and Mr. LeBeau, each of whom either had a personal stake in the procedure or was prone to potential institutional bias. Such assertions are not, alone, sufficient. Ramirez III, 810 So.2d at 844, 844 n. 13. Rather, proof of general scientific acceptance "requires the testimony of impartial experts or scientists" as to "both the underlying scientific principle and the testing procedures used to apply the principle to the facts of the case at hand." Id.at 851. D. It is clear that concern for reliability is the foundation upon which our supreme court has built its Frye test. E.g., Ramirez III, 810 So.2d at 843; Ramirez II, 651 So.2d at 1167. Thus, in Hadden, the court said: [W]e firmly hold to the principle that it is the function of the court to not permit cases to be resolved on the basis of evidence for which a predicate of reliability has not been established. Reliability is fundamental to issues involved in the admissibility of evidence.... Novel scientific evidence must ... be shown to be reliable on some basis other than simply that it is the opinion of the witness who seeks to offer the opinion. In sum, we will not permit factual issues to be resolved on the basis of opinions which have yet to achieve general acceptance in the relevant scientific community; to do otherwise would permit resolutions based upon evidence which has not *545 been demonstrated to be sufficiently reliable and would thereby cast doubt on the reliability of the factual resolutions. 690 So.2d at 578. On the record before us, the state has failed to carry its burden of demonstrating that essential portions of the expert testimony it presented at trial were based upon scientific principles and testing procedures that are generally accepted in the relevant scientific community. Accordingly, we are constrained to reverse appellant's conviction and to remand for a new trial. Should the state again attempt to establish the admissibility of the same expert testimony on remand, the trial court shall also determine whether chain of custody and contamination problems identified by appellant's experts actually exist and, if so, whether they are of sufficient magnitude to warrant their consideration on the issue of general acceptance of the testing procedures employed, rather than treating them as relevant merely to the weight of the evidence. See Murray v. State, 838 So.2d 1073, 1081 (Fla.2002). III. Appellant next contends that the trial court committed reversible error when, over a timely objection, it permitted the state to present testimony from Diane Houser that, about a month before the death of appellant's wife, Judy Ray Sybers had told her that she thought appellant was going to get a divorce. According to appellant, the out-of-court statement of one person may not be used to establish the state of mind of another. The state responds that this testimony was admissible, either because it was not offered to establish appellant's state of mind and, thus, was not hearsay; or because it constituted proper impeachment of Ms. Sybers. In the alternative, the state argues that, if the testimony was not admissible, the trial court's error in permitting it was harmless. The standard of review applicable to this issue is abuse of discretion. Ray v. State, 755 So.2d 604, 610 (Fla.2000) ("Admission of evidence is within the discretion of the trial court and will not be reversed unless there has been a clear abuse of that discretion"). However, when ruling on evidentiary matters, "a trial court's discretion is limited by the rules of evidence." Nardone v. State, 798 So.2d 870, 874 (Fla. 4th DCA 2001). We reject the state's argument that the testimony was admissible because it was offered to establish appellant's motive for committing murder and, therefore, was not hearsay because it was not being offered to prove the truth of the matter asserted. In this context, "motive" is synonymous with "emotion," "passion" or "feeling." In other words, it refers to a defendant's state of mind. See IA John Henry Wigmore, Evidence § 117, at 1697 (Tillers rev.1983). The courts of this state have repeatedly held that out-of-court statements such as that attributed by Ms. Houser to Ms. Sybers do constitute hearsay, and cannot be used to prove the state of mind or motive of someone other than the declarant. See, e.g., Woods v. State, 733 So.2d 980, 987 (Fla.1999); Kelley v. State, 543 So.2d 286, 288 (Fla. 1st DCA 1989); Fleming v. State, 457 So.2d 499, 502 (Fla. 2d DCA 1984); Bailey v. State, 419 So.2d 721, 722 (Fla. 1st DCA 1982). We also reject the state's contention that the testimony was admissible to impeach Ms. Sybers, who testified at trial that she had not spoken to appellant about divorcing his wife. As appellant correctly notes, (1) Ms. Sybers' out-of-court statement was not inconsistent with her in-court testimony because, as Ms. Houser admitted during her testimony, Ms. Sybers did not say that she had spoken with appellant *546 about divorce; and (2) Ms. Houser's testimony would not constitute proper impeachment in any event, because (if not offered to prove the truth of the matter asserted) it was designed to impeach on a collateral matter. See Wood v. State, 656 So.2d 605, 606 (Fla. 1st DCA 1995) (stating that, as a "general rule," "if a witness is cross-examined on a collateral or irrelevant matter, the cross-examiner must `take' and be bound by the answer and may not subsequently impeach the witness with extrinsic evidence in contradiction"). Accordingly, we conclude that the trial court abused its discretion when it permitted the state to introduce the testimony of Ms. Houser regarding the out-of-court statement of Ms. Sybers. Appellant also complains about the admission, again over timely objection, of the testimony of David Levin. Levin, a divorce lawyer, was permitted to testify that, under Florida law, marital assets would be divided equally on divorce and that, assuming a net worth of $5-6 million, the wife would receive between $2½-3 million. Other than the inadmissible hearsay testimony of Ms. Houser, there was no evidence that appellant had considered dissolving his marriage. Without some factual predicate, Levin's testimony was irrelevant, and should also have been excluded. Having concluded that the trial court erred when it permitted the testimony of Ms. Houser and Mr. Levin, we must address the state's harmless error argument. It is the state's burden to establish that those errors were harmless. To do this, it must demonstrate, to the exclusion of all reasonable doubt, that neither error affected the verdict. See State v. DiGuilio, 491 So.2d 1129, 1139 (Fla. 1986); Goodwin v. State, 751 So.2d 537 (Fla.1999) (applying the DiGuilio test to non-constitutional errors). Having carefully reviewed the entire trial transcript, we are unable to say, to the exclusion of all reasonable doubt, that the errors had no affect on the jury's verdict. Accordingly, the state has failed to carry its burden, and we conclude that the errors were not harmless. IV. In summary, we hold that the trial court committed reversible error when it admitted (1) the expert testimony of Dr. Ballard and Mr. LeBeau, which was based on tests purportedly establishing the presence of succinylmonocholine in the victim's embalmed tissues nine years after the victim's death; (2) the testimony of Ms. Houser that Judy Sybers told her that she thought appellant was going to try to get a divorce; and (3) the testimony of Mr. Levin regarding Florida divorce law and its impact on the distribution of marital property. Because of these errors, we reverse, and remand for a new trial. This disposition moots appellant's postconviction claims. REVERSED and REMANDED, with directions. DAVIS and VAN NORTWICK, JJ., concur.