Tyrone Jones looked like a success story. He had left his East Baltimore neighborhood to start college in Texas, he planned to marry his high school sweetheart. He had gotten out of town without a violent record.
But on a hot summer evening in 1998, minutes after 15-year-old Tyree Wright was gunned down on the steps outside an East Federal Street rowhouse, Jones -- home on break -- was arrested and charged with murder. A jury later convicted him, sending him to like in prison.
Two eyewitnesses connected Jones, then 21, to the shooting. But their testimony was flawed, and it might not have swayed the jury had it not been for prosecutors' trump card: forensic evidence.
Police had found gunshot residue particles on Jones' hand, indication, if not proof, prosecutors said, that he had fired or handled a gun.
"We couldn't have found nothing without the gunshot residue," recalled Dorothy M. Bellamy, a juror in Jones' trial. "We were told it was enough residue to bring him to court.
"Other than that, I couldn't have sat up there with good grace and say he'd done it."
But this evidence, used for years in thousands of city investigations, is not as clear-cut as it has often been portrayed in court, or in the television crime dramas that have made forensics popular.
Instead, across the country it is under increasing scrutiny as defense attorneys, prosecutors and police officials evaluate how easily suspects' hands can be contaminated by other items already covered in gunshot residue: handcuffs, car seats, even police officers themselves.
In Baltimore, prosecutors say they have begun a review of the way the evidence is used. Some agencies, such as the Boston police, have decided not to use it. Others, including the FBI and Maryland State Police, must meet a higher standard than Baltimore's before they suggest that the evidence means someone has fired a gun.
"The ability to contaminate is the reason that there is such a limited degree of conclusions that can be made with gunshot residue," said Marc S. Taylor, a gunshot-residue expert from California who has testified nationally for both prosecutors and defense attorneys. He is not connected with Jones' case. "They should never be making the statement that says gunshot residue shows that someone fired a gun."
Internal Baltimore police documents show that contamination has been a recurring problem. Despite that and the prosecutors' review, the department continues to use gunshot residue in hundreds of cases each year.
Police officials say they are confident that the evidence is solid. Procedural changes, they say, have removed any problems.
"We've pretty well covered the contamination issue as far as we're concerned," said Ed Koch, the director of the Baltimore Police Department's crime lab.
Lack of controls
Still, there are serious questions about whether it is ever possible to avoid contamination with this type of evidence, particularly when, as in the case of the Baltimore Police Department, there is no monitoring set up to identify and therefore prevent it, as would be standard in an independent laboratory.
And there has been scant effort to review older cases, no comprehensive evaluation to see whether potentially tainted forensic conclusions, used in hundreds of cases each year, have put innocent people in prison.
"People have been prosecuted and convicted on the basis of scientific evidence we know is flawed," said Patrick Kent, chief of the Maryland public defender's forensics division. "How can we make sure there's justice in our system?"
It is impossible to say whether Jones is innocent or whether contamination caused the gunshot residue findings in his case. He says he wasn't even near the shooting. The state's attorney's office says it stands by the conviction.
But many experts say no trial should turn on a tiny particle such as the one a lab technician found on Jones' left hand.
"This is an extremely vexatious case," Baltimore Circuit Judge John N. Prevas said at Jones' 1999 sentencing. "This is only one of about four in my career that I have said to the defendant, 'You know a jury of your peers has convicted you and I have to sentence you as though you did it. But if you can ever come forward with any evidence that demonstrates that you did not do it, I will be more than happy to set you free.'"
The judge continued: "You have this very strange situation where a man who was immediately identified as the defendant happened to have gunshot residue on his hands. I think that is what caused the jury to resist the fairly elaborate defense that the defendant put on as to why he was a victim of circumstances."
Gunfire in the night
The evening of June 24, 1998, wasn't the hottest night that summer. But it was late June in Baltimore, and people were out. In shorts and tank tops, they sat on steps outside sweltering rowhouses, drinking water, chatting with neighbors.
Tyree Wright and his younger brother, Emmanuel Johnson, sat in front of their mother's place, 1701 E. Federal St. It was about 10:30, and the street lamp in the narrow alley next to their house spread light over the trash and cracked concrete.
Their uncle was with them. So were their stepfather, his brother and sister-in-law. Their mother was in the house.
They were just talking, hanging, they would remember, when their uncle, David Michael Brown, yelled. He was the first to see it.
"He got a gun," Brown shouted.
Two men walked out of the alley, one aiming a silver-colored revolver at the crowd. Then came the shots. Emmanuel and his uncle jumped away from the steps and ran. Others sped away in a car.
Witnesses said the gunman was aiming at Tyree Wright.
"Me and my uncle started running down the street," Emmanuel later testified. "I turned back around to look for my brother and I seen my brother laying on the ground."
He was face down, according to witness accounts. Blood pooled beneath him. His glasses were beside his head. Police cars and flashing lights swarmed the narrow road. Tyree's mother screamed. A crowd pushed its way toward the body.
"There was a lot of chaotic behavior," Officer Donavan Cox testified.
He asked Emmanuel what had happened. The eighth-grader was frantic, the officer recalled later, even hysterical. But he responded, saying two men had come out of the alley to shoot his brother. One, he said, wore a gray tank top.
Emmanuel told the officer he didn't recall anything more. He said he hadn't seen either man's face. Cox relayed the description across his radio. Gray tank top. It wasn't much. It seemed like half the city was wearing tank tops that night.
But the police would be looking for something else when they found their suspect - gunshot residue.
Cloud of particles
Look through a microscope at the hands of someone who has just fired a gun and there will probably be hundreds of lead, barium and antimony particles. This is gunshot residue.
The same explosion that forces out the bullet also releases these particles in a fine, nearly invisible cloud. It is one of the few ways these three elements become fused.
But after a few moments, much of the residue won't be there any more. It is like talcum powder. One shake and particles scatter. One rub and they'll transfer from a gunman's hands to his pants or a car seat or even handcuffs.
That is why officers try to test a suspect for gunshot residue as soon as possible: It is easy for someone to get rid of it.
Not a 'clear result'
It's also why defense attorneys and many experts worry about contamination, and why some departments don't bother with the science. Gunshot residue, they say, can mean someone fired a gun, or was next to a gun, or touched a person who fired a gun, or touched a car seat where someone who fired a gun once sat.
"We feel that for the amount of effort you're putting into it, you're not getting a clear result back," said Elizabeth Ziolkowski, a senior criminalist with the Boston Police Department.
There are few comprehensive studies of how easily gunshot residue transfers. One internal test by the Los Angeles Police Department found that police cruisers were contaminated by gunshot residue and that the particles transferred onto people who hadn't fired anything.
Baltimore has had its contamination issues.
In 2001, the Baltimore Police Department revealed that there had been active firing ranges inside some precinct houses where suspects were tested for gunshot residue. Testing revealed gunshot residue in interview rooms, on tables, on chairs, in the air.
Suspects' hands bagged
The department says it revamped its testing process when it found out about the problem. It set up a "clean" room in a garage area outside of police headquarters for gunshot residue testing. Officers were told to bag suspects' hands at the time of arrest so that they would not be contaminated.
"If they don't bag them, we don't test them," said Koch, the crime lab director.
But it is unclear from internal police documents when this policy went into full effect or whether it has been followed consistently.
Moreover, when staff tested the "clean" room in November 2003 - part of an effort to secure accreditation for the city's police laboratory - they found gunshot residue all over the handcuffs, gun belt and holster of the officer assigned to the examination area.
The results of that test were reported internally in March 2004. Two months later, forensic supervisor Sharon Talmadge issued a mandate to her staff:
"As of todays date (5/14/2004) Police Officers assigned to the Mobile Unit WILL NOT perform GSR testing," she wrote in an e-mail.
The test results also showed dozens of particles consistent with gunshot residue on door handles, chairs and walls. A "unique" particle - the fused barium, antimony and lead distinctive of gunshot residue - was on the floor.
"You've got a contaminated room, obviously," said Taylor, the California expert.
Kent, the public defender's forensic chief, said the findings pointed to what he believes is a longstanding problem at the police lab of not having the type of control samples and monitoring system that would be standard in an independent lab.
In one of Kent's cases, quality-assurance expert Janine Arvizu wrote in an affidavit that the department's response to contamination was "ineffective" and that "in the face of objective evidence of contamination that potentially rendered GSR results invalid, BCPD made no apparent effort to assess the scope and significance of the observed contamination."
Koch said the 2004 findings did not indicate a problem.
Most of the individual particles found in the testing room were lead, which doesn't by itself indicate gunshot residue, he said. And the few "unique" particles were probably brought in on an officer's shoe and were unlikely to get onto a suspect "unless we dragged his hands across the floor," he said.
Moreover, Koch said, police are careful to explain the limits of gunshot residue evidence in trial.
"We do not indicate in court or state in court that a person shot a gun because they have gunshot residue on their hands," Koch said.
Tyrone Jones' case suggests otherwise.
Minutes after Wright's shooting, the officers noticed him. He was standing in a group near the corner of Eager and McDonogh streets, wearing a gray cutoff shirt, close enough to the description that had just come over the radio. It was about a half-mile south of the crime scene.
The officers slowed their cruiser. Later, they testified that members of the group started their unhurried, well-practiced scatter. They knew the rules about loitering, Jones explained later. To stand and stare would be asking for trouble.
One of the officers shouted at Jones to stop, according to witness testimony and Jones' recollection. They got out of the car and asked him to pull up his shirt. No gun. They told him other police would be bringing witnesses from a crime scene to check him out.
One cruiser came and left, Jones said. Then another, carrying Tyree's little brother Emmanuel, pulled up. The officers shone the car's spotlight into Jones' face.
The officer who was driving Emmanuel said at the trial that the boy never made an identification. But the officers next to Jones said they heard from the cruiser a "10-30," confirmation that Jones had been recognized as the killer. Later Emmanuel said he recognized Jones' clothing.
The officers told Jones that they were taking him to homicide.
"I'm sitting there and wondering, 'Are you serious?'" Jones recalled during a recent prison interview.
He had plane tickets to fly back to school in Texas in the morning.
Jones had never faced trouble like this. He had been arrested in Baltimore for offenses such as loitering and drug possession. But those cases were dropped or dismissed before they got to trial. That's what happened to people in his neighborhood, Jones said with a slight shrug.
His one prior offense was a drug charge in New Jersey. It was a stupid mistake, he said, transporting cocaine. It was also a wake-up call, pushing him to get out of Baltimore, take the next step with his girlfriend - who would soon be his wife - and start his computer science studies.
On the night Wright was killed, Jones testified, he had been playing basketball and had then gone with friends to CCs carryout, near Eager and McDonogh. He didn't know Tyree Wright, didn't know that neighborhood.
Two women backed him up, testifying that they had seen him in the group outside the carryout when they heard the gunshots.
Assistant State's Attorney Stephanie Royster found inconsistencies in their stories and repeatedly asked during her questioning why none of Jones' supposed basketball friends had testified.
Jones said the men who had played ball with him that night had agreed to testify for him but that his lawyer didn't want to use them because they had criminal records. His trial defense lawyer, Paul Polansky, said in a recent interview that he could not remember whether there were other possible alibi witnesses.
"But if there were, and we didn't put them on, there was a reason," Polansky said.
As the officers drove Jones to headquarters, they knew none of this. A teen was dead, and they had their suspect.
Swabbed for residue
Detectives ushered Jones into an interview room at headquarters and handcuffed him to the table, Jones recalled. They testified they called for Jawad Abdullah, a crime lab technician, and asked him to sample Jones for gunshot residue. This was before the department told officers to bag suspects' hands upon arrest.
Jones said he didn't know what the tests would mean.
"I had no idea what GSR even was," he said in a recent interview.
Abdullah testified that he dabbed around Jones' left and right hands, pressing a sticky-surfaced swab along the back of his thumb and forefinger, on webbing, along his palm.
The technician secured the dabs in two vials - one for the right hand, one for the left - and sealed them in separate bags. Those bags went into an envelope, which was sealed with yellow tape.
"And then we put it in here," Abdullah demonstrated for the jury. "Fold it, put it into another envelope that holds the evidence for Evidence Control Unit."
That's where the samples wait for gunshot residue examiners, the people who look for particles.
When those examiners looked at the evidence from Jones' case, they found no gunshot residue from the right hand, according to lab notes.
They found one unique gunshot residue particle - the fused barium, antimony and lead - on the left, Jones' non-dominant hand, according to the notes. There was also a particle of barium and antimony, which some experts call likely gunshot residue, and at least 15 lead particles, which are not considered gunshot residue particles.
At the trial, Daniel Van Gelder, one of the department's gunshot residue examiners, told the jury that police had found 17 gunshot particles on Jones' left hand.
The defense lawyer, Polansky, did not challenge him. Polansky said in a recent interview that he had not known at the trial that there was only one unique gunshot residue particle, and that Van Gelder had misstated the number.
As was standard practice at the time, the police had not given prosecutors, and prosecutors hadn't given Polansky, the lab notes underlying their report, only the final conclusion that said, "Gunshot primer residues were found on the hand(s) of the suspect."
The difference between one particle and 17 can be huge at a trial. Many experts and police departments - including the Maryland State Police - say one particle is as likely to come from contamination as from a gun.
The FBI goes further, saying one particle is not enough to label someone "positive" for gunshot residue. It needs at least three to start using the evidence, according to its protocols.
In Jones' trial, Assistant State's Attorney Stephanie Royster asked Van Gelder about the possible sources of the gunshot residue.
"Most probably it's from either firing a gun or having your hand near a gun when it went off," he responded, according to a transcript of the trial.
The prosecutor started to ask her next question when Van Gelder interrupted.
"There are other possibilities," he said. "And that is if somebody fires a gun and then hands the gun to you, that gun would contain some particles on it that would get onto your hand."
It was a particularly helpful "possibility" for the prosecution. By that point in the trial, the prosecution's theory had shifted to trying to prove that Jones was not the shooter but the shooter's accomplice.
Part of the reason was that Wright's uncle, David Michael Brown, had given a witness description that made it impossible for Jones to be the shooter.
His testimony still had flaws, even for the new theory. Brown had told police the night of the killing that he hadn't seen the face of the non-shooter but noted that he had dreadlocks, a hairstyle Jones never wore.
But Brown said he recognized Jones as one of the men who came out of the alley and picked him out of a photo lineup a week after the shooting. And, as Royster said in her closing, "every shred of scientific evidence in this case corroborates what David Michael Brown saw."
The jury acquitted Jones of all the first-degree-murder and handgun counts but convicted him of conspiracy to commit murder.
"My whole body was numb," Jones said.
At sentencing, the judge expressed his concerns but sentenced Jones to life in prison.
Reviewing closed cases
After the Baltimore Police Department's gunshot residue contamination problem became public in 2001, the defense bar clamored for action.
Baltimore State's Attorney Patricia C. Jessamy tried to find out from the Police Department how long the problem had existed, said her spokeswoman, Margaret T. Burns.
"We recognize that there needs to be some systematic method for locating and reviewing closed cases," Jessamy then wrote in a Nov. 13, 2001, letter to the city public defender. "We all need to identify which cases could be affected."
But there was never any such examination. Burns said the defense bar never responded with a group of cases it wanted reviewed. Defense attorneys say Jessamy never followed up.
"Could we have done a better job here? Yes," said Michele Nethercott, chief of the Maryland public defender's Innocence Project, who now represents Jones. "The problem, of course, is that at that time in Baltimore City, the public defender actually ended up refusing to take any more cases because of the case overload. We were barely able to deal with cases."
Asking for new trials
More recently, in part prompted by the 2004 "clean" room findings, some defendants have asked for new trials because of possible gunshot residue contamination.
One such defendant has won a new trial, but because of a legal issue involving discovery, not contamination.
Jones' motion was denied.
Nobody has been arrested as the shooter in Tyree Wright's death. Although Jones says authorities have offered him various deals, he has never given any information.
"I can't," he said. "I just kept telling them, 'I can't help you. I wasn't there.'"
Nethercott's job is to identify cases in which she believes defendants have been wrongly convicted. After sifting through reams of evidence in hundreds of cases, she takes about 20 clients a year.
As Nethercott started to look into Jones' case, she soon realized it fit in with an investigation that her colleague Kent was beginning into gunshot-residue evidence.
As they continued looking into the science, they say, the more troubled they became.
"I had no idea what I was getting into," Nethercott said.
Armed with police documents, experts who criticized the department's standards, and numerous studies about gunshot residue, Nethercott asked Judge Prevas for a new trial.
In May 2004, the judge rejected the request, calling the motion a "Hail Mary pass."
Although the decision might have seemed to conflict with the comments he made at Jones' sentencing, Prevas said his new trial decision had to be bound by whether there was actually new evidence.
There was no proof that in the room Jones was tested, on the night he was tested, there was any contamination. That part of the homicide division had been remodeled between the time Jones was arrested and the 2001 contamination study, so there were no data at all.
No new evidence
Some defense attorneys speculated that the judge was loath to expose the system to a flood of new-trial requests. But Prevas said Nethercott's suggestion that the Police Department had a widespread contamination problem, from the point of arrest on, was not enough to overturn the guilty verdict.
"There is no newly discovered evidence in this case," he said during his ruling.
Jones is appealing, and Nethercott said she is looking for other ways to challenge his conviction.
Meanwhile, Kent said, his office is trying to investigate the way gunshot residue is used in the city courts. There could be hundreds of cases in which it turned a jury, he said.
"This office will not stop until all those cases are reviewed," Kent said. "This has to be a stark wake-up call."