ninov yashar

7/30/2019 Ninov Yashar

1/35

1

Victor Ninov's 'Discovery' of Element 118:A Case of Alleged Data Fabrication

Mark Yashar

U.C. Davis, Physics 280

6/1/07

Outline

Introduction to and HistoricalBackground of Superheavy NucleiSynthesis


2/35

2

Ninov's Professional Background andevents leading up to the heavy nucleisynthesis experiments at LBNL

Description of experiments, timeline ofevents leading up to alleged misconduct

Investigations and allegations of datafabrication

Ninov's denials, reactions of co-workersand collaborators

Lessons learned, implications,conclusions, & postcripts


3/35

3

Superheavy Nuclei Synthesis:

Introduction, Motivations, and

Historical Background

Synthesis of new heavy nuclei hasfundamental interest for physics &

chemistry.

Heaviest nuclei provide lab to test

ideas of nuclear structure at limits oflarge numbers of protons and

neutrons in the nucleus.

Some of these superheavy elementshave relatively long half-lives and so

could prove useful in treating cancerand in creating medical diagnostic

procedures.


4/35

4

Periodic Table of the elements(Chemistry of Superheavy Elements by Matthias Schadel;

Angewandte Chemie International Edition, Volume 45, Issue

3, pages 368-401, 2006)


5/35

5

"classical cyclotron." Encyclopedia Britannica Online.(http://www.britannica.com/ebc/art-59676)

Three Generations of Cyclotrons(http://www.lbl.gov/nsd/user88/cycgreenbook.html)


6/35

6

88-Inch(magnetic pole diameter) Sector-Focused Cyclotron atLBNL(http://www.lbl.gov/nsd/user88/cycgreenbook.html)


7/35

7

Introduction and Historical Background

Mid-1930's: Race for elements beyond

uranium started. Involving groups inRome, Berlin, & Paris.

1940: Team at LBL used cyclotron tocreate element 93 (Neptunium) andelement 94 (Plutonium).

Neptunium: When H (element #1; 1proton in nucleus) fuses with Uranium(Z=92), produce artificial element Z=93.

Early to Mid-1940's: After furtherdiscoveries (i.e., synthesis andseparation of americium & cirium at

U.C.B.), G. T. Seaborg hypothesized thatnew series of elements called actinoidseries was being produced & that thisnew series begin with thorium (Z=90).

Thereafter, heavy-element discoverieswere sought and made in accordance

with this hypothesis.

Radical revision and expansion of theperiodic table.


8/35

8


Table of Discovery of Transuranium

elements at: (http://www.britannica.com/eb/article-9115859/Table-27-Discovery-of-the-Transuranium-Elements)

Over next 40 years, researches at LBNLdiscovered 12 more transuraniumelements, more than at any other

institution during that time.

Mid-1950s: Theoretical developmentsconcerning atomic nucleus.

Nuclear shell models/theory: similar toelectrons in atoms & molecules and

based on same quantum mechanicallaws, protons and neutrons form closed

shells with magic #s. Nuclei w/closedshells exhibit extra and sometimespronounced stability (e.g., longer half-lives).

Superheavyelements, starting withRutherfordium (Z = 104) only exist

because of their microscopic shellstabilization.


9/35

9


Island of Stability: Region of periodic

table (starting at around Z=108(N=162) Z =114 (N=184) ) consistingof superheavy elements w/half-livessignificantly longer than their slightlylighter superheavy neighbors on theperiodic table.

1961: 88 inch cyclotron at LBNL laterused by Ninov and collaborators wasconstructed.

1965 onwards (over next 25 years):researchers have sought to find orsynthesize superheavy

nuclei at or near the region Z=114 & N =184 (i.e., near center of island ofstability.

1980's: LBNL no longer world leader insynthesis of transuranium elements.

Other U.S. facilities: SLAC, BrookhavenNational Lab, etc...

Early 1990's: Changing priorities.


10/35

10

Ninov's Professional Background andevents leading up to the heavy nucleisynthesis experiments at LBNL

Byearly 1990's, GSI (Institute for HeavyIon Research) in Germany and JointInstitute for Nuclear Research, in Dubna,Russia, were taking more of leading rolein sythesizing heavy nuclei.

Early 1990's: Ninov does his PhD

(1992) and post-doc work at GSI. Mid-1990's: Ninov helps to discover

elements 110,111, and 112 at GSI withuse of data analysis code that hedeveloped. This establishes hisreputation as world-class expert in thefield.

During most of 1990's, low excitationenergy fusion reactions failed to takescientists beyond element 112-114. Extrapolations of existing data onsynthesis of elements 110-112 indicatedthat to reach still heavier elementswould require orders of magnitude

increases in accelerator beam currents &new target technologies.

Mid-Late 1990's:synthesis of elements110-114 had invogorated quest forisland of stability.


11/35

11

1996: LBL hires Ninov as part of effortto search for island of stability.

LBL research team specifically wanted

help building Berkeley gas-filledseparator (BGS) instrument, similar toone Ninov had helped to construct atGSI.

1996-1998: Ninov joined nuclearchemist and Project Leader (Kenneth E.Gregorich) on project

to construct BGS (based on designproposed by A. Ghiroso). Starting in 1996 Gregorich worked with

Ninov almost every day for next 5 years.

LBNL photo of Ninov (left) working withGregorich (right) at BGS during 1999

experiments


12/35

12

Events leading up to heavy nucleisynthesis experiments at LBNL

Ninov spent his first 2 years at LBLworking with Gregorich in constructingBGS.

Description and function of BGS:

Atoms accelerated with cyclotron and

then slammed into metal target. Newlyfused elements and other reactionproducts would then pass through BGS,consisting of magnets, He gas, anddetectors.

Strong magnets in BGS would thenfocus, sift out, and separate ions of

interest from all of un-interestingreaction products, which are produced inmuch larger quantities.

Various detectors in BGS would identifyand record energy, position, and timingof reactions associated with events,i.e., decay chains of interest.


13/35

13

Events leading up to heavy nucleisynthesis experiments at LBNL

In all work done on synthesis of heavynuclei up to that point, researchers hadfound it progressly more difficult otproduce more massive elements byfusing projectile & target.

Early 1999: Ninov & Gregorich were

finishing construction of BGS andplanning range of experiments to test,debug, improve performance, and readyit for experiments.

Approached by visiting scholar and LBLFullbright Fellow, Robert Smolanczuk,from Soltan Institute for Nuclear Studies

in Warsaw, Poland.

According to Smolanczuk'scalculations (R. Smolanczuk, Phys.Rev. C 59, 2634, 1999),under right conditions could leapacross gap of increasingly

improbable atoms elements 113-117-- and generate element 118.


14/35

14

Smolanczuk's calculations suggested

that 118 would be relatively easy tocreate by bombarding a lead target(Z=82) with a low-energy kryptonbeam (Z = 36) (i.e., chance ofproducing 118 was much greaterthan had been expected due to muchlarger nuclear reaction cross

section). Ninov & Gregorich were persuaded by

other team members to testSmolanczuk's theory.

Bypassed all testing and debuggingplans and decided to use next availablebeam-time to try to synthesize 118.


15/35

15

Description of experiments, timeline ofevents leading up to alleged misconduct

April 8-12 1999, Experiment #1April 30-May 5 1999, Experiment #2(Experiment Repeated)

Directed beam of krypton ions fromcyclotron to some very thin lead targetsmounted in B.G.S.

B.G.S. was configured to capture anddetect, with high efficiency, any ion ofelement 118 in downstream plane ofdetectors. Detectors would recordprecisely when and where ion came torest, as well as times and energies of

telltale sequences of alpha-particles (Henuclei) emitted in chain of decays tolower Z.


16/35

16

(Ninov, et al., Phys. Rev. Lett. 83, 1104 (1999))

Because of Ninovs previous work atGSI, he played pivotal role in theseexperiments.

Enormous amount of raw dataprocessed and analyzed by Ninovusing data analysis software he haddeveloped and mastered at GSI.

As only team member familiar withsoftware, Ninov was put in charge ofdata analysis.

Ninov was only one to deal withoriginal raw data, which came fromdetectors in binary form and wasstored on magnetic tape.


17/35

17

Ran computer-analysis program thatconverted binary files into text files ofwords and numbers for rest of team to

interpret.

During this time (1999), NO ONE ELSEKNEW HOW TO USE PROGRAM, SORELIED ON NINOV TO SUPPLYRESULTS.

Ninov was seeking pattern that wouldindicate Kr (36p) and Pb (82 p) hadfused to momentarily produce nucleusof 118, which would subsequentlydecay to chain of smaller elements andrelease of alpha particles.

After completing initial analysis, Ninovclaimed that he had in fact foundevidence for this decay chain.oThis decay sequence agreedparticularly well with Smolanczuk'scalculations.

oAppeared that in one instant, teamhad discovered 2 new elements(116 & 118).

oInitially, reported that 3 high-energy decay chains wereobserved (experiment #1).


18/35

18

Description of experiments,...

When Ninov showed team his dataanalysis results, everyone reactedwith surprise and caution:oExamined Ninovs results several

times.oRe-ran experiment (April 30-May

5) after making improvements toB.G.S.

oNinovs analysis of experimentalresults appeared to reveal anotherevent (decay chain) enough toconvince rest of team that theyhad real, reproducible discovery.

oAfter group had closely reviewedNinovs calculations, W. Loveland

(another member of team) filled abinder with supporting evidence; 1of decay chains was thrown out,but this still left 3 good decaychains.

oScientits at GSI also examinedresults and agreed that something

noteworthy might have occurred. Everyone was working from numbers

Ninov had obtained from his analysis.No one felt need to examine originalraw data.


19/35

19

Description of experiments,...

Pressure/Competition from other labsin, e.g., Germany and Russia.

Researchers quickly wrote up theresults and submitted them to PhysicalReview Letters.

August 9, 1999: Paper published inP.R.L. on synthesis of element 118.


20/35

20

(LBNL; Physic Today.org, A.I.P. 2003,

http://www.aip.org/pt/vol-55/iss-9/p15.html)

(LBL Research Review, Summer 1999,http://www.lbl.gov/Science-Articles/Research-

Review/Magazine/1999/departments/breaking_news.shtml)

Gregorich: During 11 days ofexperiments, 3 alpha decay chains were

observed, indicating production of 3

atoms of 118 Decay energies andlifetimes for these new isotopes ofelements 118,116,114, 112,110,108, &106 provide strong support for existenceof predicted island of stability


21/35

21

Investigations and Allegations of DataFabrication

Summer 1999 GSI team (lead by S. Hofmann) repeated

LBL experiment with its own cyclotronbut did not detect any 118 decay chains.

Research groups in Japan and Francealso repeated experiment, but did notfind any evidence for 118.

Spring 2000 Ninov, Gregorich, and rest of LBL

repeated experiment, but found no traceof experiment.

Possible Explanations:

Statistical variations due to randomprocesses?

Computer program playing tricks onresearch team?Internal independent review committeewas assembled to investigate.

Made several recommendations toimprove on original experiment.oSuggestion: Have several people

analyze raw data. (But thissuggestion was not taken upimmediately).


22/35

22

Investigations and Allegations

oRuled out most obvious technicalexplanations:

discrepancies in beam alignment detector inefficiencies flaws in gathering & processing

data.

No consideration of possible fraud atthis point.

April 2001Improvements made in detection

equipment and Gregorich & Ninov againgot beam time to retry experiment.

oInitially, Ninov reported that he hadfound another 118 decay chain

By now, W. Loveland had learned how touse Ninovs data analysis software, anddid not find decay chain reported byNinov in his re-analysis.

Ninov and several other collaboratorsalso re-analyzed the data with hissoftware and did not find decay chain.

Colleagues finally re-analyized 1999experiment results, some using theirown data analysis programs, and foundthatthere were no decay chain events inthe original (raw) data files.


23/35

23


By this time, LBL had appointed 2ndinternal review committee whichincluded computer experts.oAppeared as if the 3 events reported

in 1999 were in processed text filethat several researchers hadexamined, butNOTin original binaryfile.

3rd review committee, including workinggroup chaired by D. Hoffman, seniormember of B.G.S. team, examined both99 and 01 data files, and found thatnone of them contained record of anydecay sequence (proof) for elements116 or 118.

Plausible theory: someone had insertedfalse data into text file , and Ninovmight have been in best position to dothis (?)

Issue needed to be examined by expertsfrom outside the research team.

June-July 2001

1. Gregorich & colleagues issuedpress release withdrawing discoveryof element 118 and told researchersat other labs to disregard claim forelement 118.


24/35

24

Investigations and Allegations RETRACTION

July 26, team submitted letter to PRLretracting its original finding.

o1 short paragraph offering fewdetails, letter simply states thatexperiment had not detected 118 in1999.

Ninov stood by original findings andrefused to sign retraction letter, so PRLjournal editors rejected retraction letter.

Ninov had insisted to journal editorsthat it was premature to repudiatediscovery before more experimentswere done, and this delayed publicationof retraction later until July 2002.


25/35

25


oNinov later complained thatretraction had been submittedbehind his back.

Fall 2001:

Head of LBNLs computing division,Stewart Loken, had received technicalreport from yet another review committee

and had conducted interviews with Ninovand other principals.

oRecommendation: Lab shouldproceed to formal investigation ofNinovs conduct under LBNLs statedpolicy on integrity in research.

November 2001

LBL puts Ninov on paidleave onNovember 21. Lee Schroeder, director of LBNLs

nuclear science division team, convenesformal investigative committee chairedby Caltech physicist Rochus Vogt (VogtCommittee) a week after Ninov place on

administrative leave.


26/35

26

December 2001

In process of writing up paper ondiscovery of element 112, Hofmann and

colleagues at GSI went back to originaldata tapes and re-analyzed data fromoriginal 1994-1996 experiments onelements 110-112 that Ninov wasinvolved with.oFound that 2 decay chains reported

in original papers were not in

original raw data files.oDecay chains appeared only in text

files that Ninov had produced inprimary data.

o Given that Ninov was in charge ofdata processing and analysis at thattime, Hofmann believes that

evidence points squarely at Ninov forcreating these false events.

oPossible evidence here that Ninovmay have been fabricating alpha-decay chains since 1994.

March 2002

Taking this and all other facts intoaccount Vogt committee reported itsfindings to Schroeder and LBNL directorCharles Shank at LBNL:


27/35

27

We find clear and convincing evidencethat data in 1999, upon which reporteddiscovery was based, were fabricated

[We] regret that our findings revealedintentional fabrication instead of honesterror or honest differences ininterpretation.

.. There is clear evidence that Ninov

[carried] out this fabrication If anyoneelse had done [it], Ninov would almostsurely have detected it.

Possible Smoking Gun: computer logfile created by Ninovs data analysissoftware that had automatically

recorded everything that had occurredduring handling of 1999 and2001 runs.opage lengths were inconsistentotiming of some events was offoelements passed off as 118 decay

chain could have been manufacturedby cutting and pasting few lines from

elsewhere in file and changing somenumbers , i.e., some of originalchains may have been edited bysomeone using account Vninov .


28/35

28


Vogt report closes with criticism ofteams procedures in 1999 experimentandcriticism of co-authors andcollaborators:

[We] find it incredible that no one else ingroup, other than Ninov, examinedoriginal data to confirm purported

discovery of 118 Responsibility of co-authors.

May 2002LBNL fires Ninov

July 2002

Official retraction finally published in PRLas Editorial Note with introductoryexplanation pointing out that all but oneof authors of original Letter have asked usto publish following retraction

August 2002

LBNL makes Vogt report public.


29/35

29

Ninov's denials, reactions of co-workersand collaborators

Ninov emphatically denied accusationsof Vogt Committee. Fall 2002-- initiated grievanceproceedings and prepared detailedrebuttal of charges, which he claimedwere filled with errors andcontradictions

Notes that there was simply nomotivation for scientist with hisimpressive publication record to commitfraud and in such sloppy manner.

Acknowledges that decay chains are notin raw data and admits that data filesappear to have been tampered with.

Perplexed as anyone. Computer account on lab computersystem was used by everyone in hisgroup and his password was opensecret.

Contends that any colleague could havecarried out deception.


30/35

30

Why create data so bad that flaws canbe detected in few min. of examination? Why did my expert colleagues never

question obviously flawed data? Why,having apparently successfullyperpetrated scientific fraud, did I neverthink to delete the incriminatingevidence ..? .. To these questions,answer can only be because the file wasnot of my creation Im a scapegoat

Conspiracy theories?Reaction of Colleagues

Perplexed why would promisingscientist decide to fabricate data?

D.M. Lee: Its unbelievable that anyonewould do this sort of thing There wasabsolutely no need for him to do this

his career did not depend on this. Ghiorso: speculates that Ninov was

buying time for team by inserting eventsso that lab would let experiment runlonger and provide opportunity to catchsome real decay chains.

Gregorich and other collaborators feltthat criticism of Vogt committee directagainst team was unjustified, arguingthat no-one operates on suppositionthat his/her collaborator is fabricatingdata


31/35

31

Lessons learned, implications,conclusions, & postcripts

LBNL officials maintain that severalpeople should have been involved inessential step of data analysis, but otherscientists say even that safeguard cantalways thwart fraud.

Prof. Denis L. Rousseau: In science, wealways depend on integrity of co-

workers. When that breaks down, itsvery difficult to correct for that

oScience has built in protection,however, in fundamental rule thatinvestigators must try to reproduceinteresting findings

oNecessary to repeat other peoplesexperiments, not just for catchingcases of fraud, but for catchinggenuine mistakes as well.

LBNL Deputy Director Pier Oddone(2002): The message is thatexperimenters must exercise vigilance,

not so much against fabrication, butagainst honest error and misjudgement.And against these, first line of defenseis vigorous, independent checkingwithin experimental group


32/35

32

November 2002As direct result of Schon and Ninovincidences, APS issued new ethics

guidelines:

APS Expands and Updates Ethics andProfessional Conduct Guidelines forPhysicistshttp://www.aps.org/publications/apsnews/200301/guidelines.cfm

APS Expands and Updates Ethics and ProfessionalConduct Guidelines for Physicists

Prompted by recent highly publicized episodes

of misconduct in physics, the APS has updated

and expanded its professional ethics guidelines.

The changes, adopted November 10, 2002, at the

APS Council meeting, clarify the roles andresponsibilities of coauthors

The APS Council has adopted new Guidelines on theResponsibilities of Coauthors and Collaborators. Theguidelines state that "all coauthors share some degree ofresponsibility for any paper they coauthor" and that"some coauthors have responsibility for the entire paper.

These include, for example, coauthors who areaccountable for the integrity of the critical data reportedin the paper, carry out the analysis

To assist coauthors in fulfilling their responsibilities, theAPS Council resolved that, "Collaborations are expected to


33/35

33

have a process to archive and verify the research record;to facilitate internal communication and allow authors to befully aware of the entire work; and respond to questionsconcerning the joint work and enable other responsiblescientists to share the data. All members of a collaborationshould be familiar with, and understand, the process."

"These actions are our initial response to the recentfindings of major research misconduct"

June 2003

Goerge Trilling Co-authors areresponsible too, Physics World, June2003 (http://physicsworldarchive.iop.org/)


34/35

34

Postscripts

Ninovs denials refuted and not takenvery seriously by LBNL.

Grievance hasnt gone anywhere. 2004-?: Appealed LBNL firing decision

(?)

2003, 2004 - ?: Teaches undergradphysics at UOP, where his wife wasprofessor while he worked at LBNL.

October 2006: LLNL and Joint Institutefor Nulcear Research (Dubna, Russia)report discovery of element 118, usingapproach different from that proposedby Smolanczuk. Results published in

October 2006 edition of Physical ReviewC. Progress towards mapping island ofstability.

Additional Key References: Atomic Lies How one physicist may

have cheated in the race to find newelements, Richard Monastersky,

Chronicle of Higher Education, August16, 2002http://chemed.chem.pitt.edu/joeg/documents/chronicle_16Aug02.pdf

At Lawrence Berkeley, Physicists Say aColleague Took Them for a Ride,George Johnson, N.Y. Times, October 15


35/35

35

2002 (http://sanacacio.net/118_saga/story.html)

ninov yashar

Documents