MEMORANDUM OPINION AND ORDER
BABCOCK, District Judge.
Defendants move to dismiss this action contending that the definition of a controlled substance analogue as applied here is unconstitutionally vague. 21 U.S.C. § 802(32)(A). Hearing on the motion was held on November 19, 1992. Because the definition of “analogue” as applied here provides neither fair warning nor effective safeguards against arbitrary enforcement, it is void for vagueness. Defendants’ motion is granted and this action is dismissed.
Defendants are charged in a four cotint indictment with distribution of alphaethyl-tryptamine (AET), in violation of 21 U.S.C. §§ 813, 841, and 846. AET is not a scheduled controlled substance. Rather, .the indictment alleges that it is a “controlled substance analogue” having a substantially similar chemical structure to dimethyltryp-tamine (DMT) and diethyltryptamine (DET), both schedule I controlled substances. This is the first case where the government has prosecuted the distribution of AET under the controlled substance analogue statute.
Unless otherwise noted, the unique operative facts underlying this prosecution are undisputed. AET is an anti-depressant developed and marketed in 1960 by the Upjohn Chemical Company as “Monase.” After several years it was discovered to have toxic side effects in some patients and it was taken off the market. However, to this day, the public can purchase AET without any restrictions directly from two chemical manufacturers. Defendants here allegedly purchased AET from Sigma Chemical Company and it was allegedly delivered to. them through the U.S. mails.
There is great diversity of opinion whether AET’s chemical structure is substantially similar to DMT or DET so as to constitute a controlled substance analogue.
See, infra,
Section I. Indeed, Robert K. Sanger, chief of DEA’s western field laboratory, stated in a 1990 DEA memorandum “there is a great diversity of opinion whether this material is controlled as an analogue under the 1986 Act.”
Defendants presented the affidavit of Drs. James Ruth and Charles Duncan, who are both Ph.D. neuropharmacologists at the University of Colorado School of Pharmacy. Dr. Ruth also testified at the hearing on the motion. These experts conclude that AET is not substantially similar to DMT or DET, and that AET may not be derived by minor manipulations or tinkering with the DMT or DET molecule. Further, they conclude that AET does not have a hallucinogenic or stimulant effect on the central nervous system that is substantially similar to DMT or DET. Ruth testified that the mechanism through which AET effects the central nervous system is different than the mechanism of hallueinogen-ics and stimulants. Finally, they believe that other scientists in their field would agree with their conclusions.
The government presented a contrary scientific opinion. Drug Enforcement Administration (DEA) chemist Frank Sapienza testified that AET is a controlled substance analogue that has a substantially similar
chemical structure. Sapienza agrees that AET cannot be synthesized from either the DMT or the DET molecule. Rather, he concludes that AET is substantially similar to DMT and DET because they all share a structural family root, the tryptamine family, and they all produce some degree of hallucinogenic and stimulant activity. However, another DEA chemist, Roger Ely, testified at the hearing that AET’s chemical structure is not substantially similar to DMT or DET. Ely focused on another part of the AET molecule, the amine group, and noted that AET is a primary amine while DMT and DET are tertiary amines. Sapienza concluded that reputable scientists in this field disagree even on the methodology applicable to determine structural similarity.
Defendant Forbes was arrested by Boulder, Colorado police in 1990 for the distribution of AET, among other drugs. Because Colorado did not have an analogue statute, the Boulder County district attorney referred the prosecution of the AET sale to the United States Attorney’s office. Although subject to dispute, I find from the evidence that assistant U.S. attorney Ken Buck declined to prosecute because he determined that AET was not a controlled substance analogue. I further find from the evidence that, through discovery in his state case, Forbes obtained a copy of the DEA report stating that Buck declined to prosecute due to the conflict within the government as to AET’s structural similarity to DMT and DET. He also obtained a copy of Ely’s report on which Buck’s decision was based.
Defendants now move to dismiss this prosecution as violating the due process clause of the Fifth Amendment to the United States Constitution. They argue that the statutory definition of a controlled substance analogue as applied to AET under the circumstances of this case is unconstitutionally vague. Because this argument is premised upon a disputed question of statutory construction, I will address that issue first and then turn to the vagueness question.
I.
Defendants are charged in this case with violations of 21 U.S.C. § 813 which provides that “A controlled substance analogue shall, to the extent intended for human consumption, be treated ... as a controlled substance in schedule I.” The statute defines controlled substance analogue as:
(32)(A) Except as provided in subpara-graph (B), the term “controlled substance analogue” means a substance—
(i) the
chemical structure of which is substantially similar to the chemical structure of a controlled substance in schedule I or II;
(ii)
which
has a stimulant, depressant, or hallucinogenic effect on the central nervous system that is substantially similar to or greater than [such effect] of a controlled substance in schedule I or II; or
(iii) with respect to a particular person,
which
such person represents or intends to have a stimulant, depressant, or hallucinogenic effect on the central nervous system that is substantially similar to or greater than [such effect] of a controlled substance in schedule I or II.
21 U.S.C. § 802(32), (emphasis added).
Defendants contend that this section requires a two-pronged definition. The first prong requires a substantially similar chemical structure. The second prong requires either a substantially similar effect on the human nervous system or the intent to have such an effect. The government argues that a substance may be an analogue if it satisfies any of the three clauses. I agree with defendants.
The goal of statutory construction is to effectuate the intent and purpose of Congress.
Rocky Mountain Oil & Gas Assoc. v. Watt,
696 F.2d 734, 745 (10th Cir.1982). Where the language of the statute itself is unambiguous, it is presumed that it expresses congressional intent and the language is controlling.
Id.
If the statutory language is unclear after resort to traditional tools of statutory construction, courts look to the legislative history to glean the intent and purpose of Congress.
Blum v. Stenson,
465 U.S. 886, 896, 104 S.Ct. 1541, 1547, 79 L.Ed.2d 891 (1984);
Rocky Mountain Oil,
696 F.2d at 745.
The government’s reading of the analogue definition has superficial appeal. As a matter of simple grammar, when an “or” is placed before the last term in a series, each term in the series is usually intended to be disjunctive. Under this reading, AET would be an analogue if it satisfies any of the three clauses. However, this reading ignores other grammatical principles that apply in favor of defendants’ construction. The operative segments of clauses (ii) and (iii) both begin with the word “which”, signaling the start of a dependent relative clause modifying a precedent noun. In each case, the precedent noun is “chemical structure” found in clause (i). Because both clauses (ii) and (iii) can be read to modify clause (i), the statutory language can be fairly read as requiring the two-pronged definition asserted by defendants.
See, First Charter Financial Corp. v. U.S.,
669 F.2d 1342, 1350 (9th Cir.1982), (Modifying phrases generally refer to immediately preceding phrase). At best, the statute is ambiguous.
Defendants’ reading is also bolstered by a deeply rooted rule of statutory construction. A statute must be construed to avoid unintended or absurd results.
American Tobacco Co. v. Patterson,
456 U.S. 63, 71, 102 S.Ct. 1534, 1538, 71 L.Ed.2d 748 (1982);
Ewing v. Rodgers,
826 F.2d 967, 970 (10th Cir.1987). If I adopt the government’s construction and read clause (ii) independently, alcohol or caffeine would be controlled substance analogues because, in concentrated form, they can have depressant or stimulant effects substantially similar to a controlled substance. Likewise, if I read clause (iii) independently, powdered sugar would be an analogue if a defendant represented that it was cocaine, effectively converting this law into a counterfeit drug statute. In both cases, a defendant could be prosecuted for selling a controlled substance analogue even though the alleged analogue did not have chemical structure substantially similar to a schedule I or II controlled substance. Therefore, to prevent this unintended result, clause (i) must apply to any substance that the government contends is a controlled substance analogue.
Further, defendants’ construction is supported by legislative history. In July, 1985, the Senate began consideration of the “Designer Drug Enforcement Act of 1985” (S.1417, later redesignated the “Controlled Substance Analog Enforcement Act of 1985”, S.1437). The bill’s stated purpose was “to prohibit persons who specifically set out to manufacture or to distribute drugs which are substantially similar to the most dangerous controlled substances from engaging in this activity.” S.Rep. No. 196, 99th Cong., 1st Sess. 5 (1985). The Senate Judiciary Committee reported that law enforcement authorities find themselves one step behind underground chemists who slightly alter the molecular structure of controlled substances to create new drugs.
Id.
at 1-2. The Senate proposed a two-part definition of the term analogue: either the substance has a substantially similar chemical structure or it was “specifically designed” to produce an effect substantially similar to schedule I or II drugs.
In July, 1986, the House of Representatives considered the Designer Drug Enforcement Act of 1986 (H.R. 5246). As with the Senate, the House bill focused on underground chemists who seek to evade the drug laws by slightly altering a controlled substance. H.R.Rep. No. 948, 99th Cong., 2d Sess. 4 (1986). The House proposed a two-pronged definition of analogue that is virtually identical to the construction advocated by defendants here. The House bill contained the same three clauses as the current statute, but added the word “and” after clause (i).
The term “controlled substance analogue” is defined to conform as closely as possible to the policy of the Controlled Substances Act by
requiring a chemical relationship to a substance which is controlled ... and either
the existence of some stimulant, depressant, or hallucinogenic effect on the central nervous system, or a representation or intent that the substance have a stimulant, depressant, or hallucinogenic effect substantial
ly similar to, or greater than, such effect of any controlled substance.
Id.
at 2, (emphasis added). The report illustrates the point with coffee. “Coffee, for example, has a stimulant effect on the central nervous system, but it is not chemically substantially similar to' a controlled substance.”
Id.
at 7.
Congress ultimately adopted the analogue statute as part of the comprehensive “Anti-Drug Abuse Act of 1986” (H.R. 5484). Inexplicably, the analogue definition enacted by Congress dropped the word “and” after clause (i). Otherwise, that definition and the two-pronged definition considered by the House are virtually identical. I consider the House report to be persuasive legislative history as to Congress’ intent underlying its definition of a controlled substance analogue.
The legislative history thus clearly supports defendants’ construction. The analogue statute is directed at underground chemists who tinker with the molecular structure of controlled substances to create new drugs that are not scheduled. If a substance could be an analogue without a substantially similar chemical structure, Congress’ stated purpose would be significantly expanded. Moreover, by essentially adopting the House definition, Congress evidenced its intent to require a two-pronged definition.
Additionally, the government and defense experts who testified agree that the question of structural similarity must, in part, be evaluated in conjunction with the molecule’s hallucinogenic and stimulant activity. Because structurally similar substances have similar pharmacological effects on the central nervous system, a finding of such similar effects is some indication that the molecular structures should be classified as substantially similar. The scientific interdependence of molecular structure and effect on the central nervous system is consonant with the legislative history evidencing congressional intent to establish a dependent, two-prong test.
Finally, courts must construe criminal statutes narrowly in favor of lenity to the accused.
United States v. Enmons,
410 U.S. 396, 411, 93 S.Ct. 1007, 1015, 35 L.Ed.2d 379 (1973);
Mahn v. Gunter,
978 F.2d 599 (10th Cir.1992), (Rule of lenity is a doctrine of last resort in statutory construction). Although it is not necessary to apply this rule as a last resort, its application yields a result entirely consistent with defendants’ construction. The two-pronged definition not only promotes lenity in this prosecution, but narrows future analogue prosecutions to only those cases where the substance has a chemical structure substantially similar to a schedule I or II controlled substance.
Therefore, I hold that a substance may be a controlled substance analogue only if it satisfies clause (i)
and
clauses (ii)
or
(iii).
II.
Defendants argue that the statutory definition of controlled substance analogue is unconstitutionally vague as applied to AET under the circumstances of this case. While not contesting defendants’ characterization of scientific opinion, the government argues that whether AET is a controlled substance analogue is a question for the trier of fact. It further argues that the phrase “substantially similar” is not unconstitutionally vague.
The void for vagueness doctrine stems from the Fifth Amendment’s constitutional guarantee of due process.
Connally v. General Constr. Co.,
269 U.S. 385, 391, 46 S.Ct. 126, 127, 70 L.Ed. 322 (1926). “[A] statute which either forbids or requires the doing of an act in terms so vague that men of common intelligence must necessarily guess at its meaning and differ as to its application violates the first essential of due process of law.”
Id. See also, Hejira Corp. v. MacFarlane,
660 F.2d 1356, 1365 (10th Cir.1981), (The words of a penal statute must “be clear and beyond speculation”).
Vague laws offend several important values. First, because we assume that man is free to steer between lawful and unlawful conduct, we insist that the laws give the person of ordinary intelligence a reasonable opportunity to know what is
prohibited, so that he may act accordingly. Vague laws may trap the innocent by not providing fair warning. Second, if arbitrary and discriminatory enforcement is to be prevented, laws must provide explicit standards for those who apply them. A vague law impermissibly delegates basic policy matters to policemen, judges, and juries for resolution on an ad hoc and subjective basis, with the attendant dangers of arbitrary and discriminatory applications, (footnotes and citations omitted).
Grayned v. City of Rockford,
408 U.S. 104, 108-109, 92 S.Ct. 2294, 2298-99, 33 L.Ed.2d 222 (1972). Thus, a law is unconstitutionally vague if it fails to give fair warning of prohibited conduct or if it allows arbitrary enforcement. A statute may be vague not only on its face, but also as applied to a particular circumstance.
U.S. v. Schneiderman,
968 F.2d 1564, 1568 (2nd Cir.1992);
U.S. v. Protex Industries, Inc.,
874 F.2d 740, 743 (10th Cir.1989).
Defendants contrast this case with two recent decisions which together hold that a statute is not vague if the meaning of the statutory standard is generally accepted when applied to a particular situation. In
U.S. v. Jackson,
968 F.2d 158, 161 (2d Cir.1992),
petition for cert. denied,
— U.S. —, 113 S.Ct. 664, 121 L.Ed.2d 589 (1992), the Court examined whether the term “cocaine base” was unconstitutionally vague. The court held that, where Congress uses a technical word or term of art, it is proper to explain such term by reference to the relevant art or science.
Id.
Because cocaine base has a generally accepted meaning in the scientific community of which the defendant could have been aware, the court held that the law was not void for vagueness.
Id.
at 162-63.
In
U.S. v. Vasarajs,
908 F.2d 443, 448-49 (9th Cir.1990), the defendant was charged with reentering a military reservation after being permanently excluded. She challenged that statute as impermissibly vague because the boundary of the reservation was not adequately marked. The court stated: “[D]ue process requires that there have been some way for [defendant] to learn the boundary of the Fort.”
Id.
at 449. The court further noted that “The Constitution is satisfied if the necessary information is reasonably obtainable by the public.”
Id.
Because the defendant there could determine the geographical boundaries of the reservation, the court denied her vagueness challenge.
Here, the relevant statutory standard is the phrase “the chemical structure of which is substantially similar to the chemical structure of a controlled substance in schedule I or II.” Clearly, this standard incorporates a scientific term of art and requires reference to the appropriate fields of chemistry and pharmacology to determine its meaning. Although the “substantially similar” language may be generally susceptible to adequate definition, it runs afoul of the vagueness doctrine when it is applied to AET under the circumstances of this case.
It is undisputed that there is no scientific consensus whether AET has a chemical structure that is substantially similar to DMT or DET. The government’s own chemists cannot agree on this point, and the U.S. attorneys’ office once before declined to prosecute defendant Forbes for the very conduct that is charged here. The. scientific community cannot even agree on a methodology to use to determine structural similarity. Thus, unlike the meaning of cocaine base or the boundaries of a military reservation, a defendant cannot determine in advance of his contemplated conduct whether AET is or is not substantially similar to a controlled substance.
See, Gentile v. State Bar of Nevada,
— U.S. —, — - —, 111 S.Ct. 2720, 2722, 115 L.Ed.2d 888, 906-907 (1991), (“In the context before us, these terms have no settled usage or tradition of interpretation, in law”). Indeed, based on the prior refusal to prosecute Forbes, these defendants believed that AET was not a controlled substance analogue. To this day, AET is freely available for purchase through the U.S. mails, and the government has not scheduled it as a controlled substance throughout the drug’s 30 year existence.
Moreover, the definition of controlled substance analogue does not require any scienter—a defendant does not have to “know” that a substance has a substantially similar chemical structure to an illegal drug.
See, The Village of Hoffman Estates v. The Flipside,
455 U.S. 489, 499, 102 S.Ct. 1186, 1193, 71 L.Ed.2d 362 (1982), (A law's vagueness may be ameliorated if it requires proof of knowledge). Thus, in cases such as this one where there is no scientific consensus, criminal culpability will turn solely on a “battle of the experts” at trial. Therefore, I conclude that a person of reasonable intelligence faced with these facts must necessarily guess whether AET is a controlled substance analogue.
This conclusion is buttressed by the announced purpose of the controlled substance analogue statute. Congress declared that the purpose of the statute is to attack underground chemists who tinker with the molecules of controlled substances to create new drugs that are not yet illegal. None of the defendants here engaged in such conduct. They are not chemists who created or marketed a new designer drug. Rather, they allegedly purchased and distributed a substance that
pre-existed
the drugs to which it is a purported analogue. Under these circumstances, defendants had no fair warning that their alleged conduct was proscribed.
Nevertheless, the government argues that whether AET is a controlled substance analogue is a question of fact for the jury that cannot be resolved by the court before trial. Of course, whether AET is a controlled substance analogue would ultimately be a question for the trier of fact. However, the question posed by defendants’ motion is not whether AET is an analogue of DMT or DET—the dispositive issue is whether the statutory definition as applied to AET under the circumstances here is unconstitutionally vague. Next, the government argues that the phrase substantially similar is not vague and has been upheld in other contexts. Again, the government misstates the relevant question. While substantially similar may be adequate to pass constitutional muster in other contexts, it does not when applied to AET under the circumstances of this case.
Further, the government’s reliance on
United States v. Granberry,
916 F.2d 1008 (5th Cir.1990) and
United States v. Desurra,
865 F.2d 651 (5th Cir.1989) is misplaced. Although both cases held that the analogue statute is not facially vague, neither decision addressed the precise question presented here—whether the definition of a controlled substance analogue is unconstitutionally vague as applied to AET under circumstances substantially similar to those here. Thus, they do not compel a result or speak with any persuasive authority. Indeed,
Granherry
bolsters defendants’ argument that they had no fair warning that their conduct involving AET was covered by the analogue statute.
[T]he term “controlled substance analogue” in § 813 is clearly and specifically defined.... It provides adequate notice of what conduct is prohibited. The statute makes plain that drugs
which have been chemically designed to be similar to controlled substances,
but which are not themselves listed on the controlled substance schedules, will nonetheless be considered as schedule I substances....
Granberry,
916 F.2d at 1010, (emphasis added). Upjohn developed AET in 1960 as a prescription anti-depressant. Obviously, it was not “chemically designed to be similar to controlled substances.”
Lastly, the government relies on recent Tenth Circuit law holding “the mere fact that different minds may reach different results when seeking to determine whether a given object falls within the statutory definition of drug paraphernalia does not render the statute void for vagueness.”
U.S. v. Murphy,
977 F.2d 503, 506 (10th Cir.1992). The government’s reliance on this case is misplaced. In contrast with the drug paraphernalia cases, this action involves the application of the analogue definition to only one substance. As applied to AET under these circumstances, there simply is no core meaning to the statutory definition.
Perhaps more importantly in this case, the analogue definition as applied to AET
is so vague as to permit arbitrary enforcement. In 1990, the DEA investigated defendant Forbes for allegedly distributing AET, but the U.S. attorneys’ office declined to prosecute, citing DEA chemist Roger Ely’s conclusion that AET is not substantially similar to DMT or DET. Now, in 1992, Forbes is prosecuted by the same office for the same alleged offense. Nothing changed in the intervening two years except the personalities of the government prosecutors and their hand-picked DEA chemists.
This prosecution illustrates precisely the evils attending delegation of basic policy decisions for ad hoc, subjective resolution by those who wield prosecutorial power.
Grayned,
408 U.S. at 109, 92 S.Ct. at 2299. The government has not scheduled AET as a controlled substance over the last 30 years and it has not prosecuted the chemical company that manufactured the substance and sold it to these defendants. The government presents no non-arbitrary reason why it has now decided to bring the full weight of the criminal process to bear on these defendants. Moreover, the statute has no specific standards channeling prosecutorial discretion, thus distinguishing this case from the drug paraphernalia cases.
See, e.g., United States v. 3250 Brighton Blvd., 785
F.Supp. 141 (D.Colo.1992).
Although I recognize the strong presumption favoring constitutionality of statutes and requiring courts to construe statutes in a constitutional manner, there is no construction possible here that would satisfy due process without re-writing the statute.
See, Murphy, supra,
at 505. Thus, for the reasons set out above, I conclude that defendants have clearly and convincingly rebutted the presumption of constitutionality in this case.
I hold that the definition of controlled substance analogue as applied to AET under the unique facts here is unconstitutionally vague. Without doubt, it provides neither fair warning nor effective safeguards against arbitrary enforcement.
Accordingly, IT IS ORDERED THAT:
(1) Defendants’ motion to dismiss is GRANTED; and,
(2) This action shall be DISMISSED.