Page 1

Volume 15, Number 2 ����� Fall 1999

����� Research Report

M. Dresselhaus, K. Kneipp,S.Brown, P. Corio, A. Marucci,M.Pimenta and G. Dresselhaus

Figure 1: Structure of an armchair carbon nanotube.

����� Research Report

Optical Spectroscopy toDiagnose Alzheimer’sDisease

Gene Hanlon, G. R. HarrisionSpectroscopy Laboratory

continued on page 2

The authors are with MIT, MIT LincolnLaboratory, and The Technical Universityof Berlin

This is a collaborative project with theVeteran’s Affairs Hospital in Bedford, MA.Collaborators include R. Dasari, M. Feld, R.Ferrante, I. Itzkan, N. Kowall, D. Lathi, andA. McKee

continued on page 5

Resonance RamanSpectroscopy of Nanotubes

Single wall carbon nanotubes (SWNT’s)are cylindrical nanostructures, one carbonatom thick, about 20-50 carbon atoms aroundthe circumference, and several microns long( Fig. 1). They have remarkable electronicproperties, insofar as they can be either me-tallic or semiconducting, depending on thenanotube diameter and orientation of the

hexagons of the structure with respect to thenanotube axis. SWNT’s also exhibit re-markable Raman spectra, which can providea wealth of information about the one dimen-sional (1-D) electronic density of statesthrough a strong resonant coupling betweenthe incident and scattered photons and theelectronic transitions between the van Hovesingularities in the 1-D density of states inthe valence and conduction bands. Throughthe wide range of laser frequencies availablein the Raman Laboratory of the Spectrosco-py Laboratory, it has been possible to carryout resonance Raman experiments at manylaser excitation energies, thereby allowingstudy of many novel features.

Figure 2 shows the Raman spectrum of acarbon nanotube sample over a wide fre-quency range. The strong feature at ~155cm-1 is associated with the radial breathingmode, in which every carbon atom vibratesin phase in the radial direction. Since theradial breathing frequency is proportional to

����� ALSO IN THIS ISSUE…

����� Personalities :Mildred S.Dresselhaus

����� November 5th Workshop:Oncology and Spectroscopy

����� Fall Seminar Series:Modern Optics and Spec-troscopy

����� Recent SpectroscopyLaboratory Publications

Alzheimer’s disease (AD) is a terminalform of dementia resulting from progressivedegeneration of the neurons. Its cause isunknown. AD is the most common demen-tia and its social and public health impact isexpanding with the expanding elderly per-centage of our population. At present,definitive diagnosis is only possible by his-

tologic examination of brain tissueunder the microscope, a procedureconducted almost exclusively at au-topsy. Lack of an accessible,non-invasive technique for diag-nosing AD is the greatestimpediment in the search for itstreatment and prevention.

Recent results in the Spectrosco-py Laboratory’s Laser BiomedicalResearch Center, in collaborationwith the NIH-funded Boston Uni-versity Alzheimer’s DiseaseResearch Center at Bedford Veter-ans Affairs Hospital, indicate thatnear-infrared (NIR) Raman and flu-orescence spectroscopy candifferentiate the brain tissue of ADpatients from that of normal, agedpatients, and therefore may provide

Figure 2: First and second-order Raman spectra for carbonnanotubes over the broad phonon frequency range 100-3700cm-1 for E

laser = 1.96, 2.19, and 2.41 eV.[1]

Page 2

Resonance...continued from page 1the inverse nanotube diame-ter, the Raman effect can beused as a sensitive structuralprobe of nanotube samples,which have diameters on theorder of ~ 1 nm. At a givenlaser excitation energy, onlythe nanotube interband transi-tions between singularities inthe 1-D density of electronicstates that are resonantly ex-cited will contribute stronglyto the Raman signal (Fig. 3).Thus, by measuring the radialbreathing mode spectra formany values of laser energy,E

laser, we can calibrate the di-

ameter distribution of ournanotube samples.

The major focus of our re-search at the SpectroscopyLaboratory has been directedto the study of tangentialcarbon-carbon stretch modesin the 1500-600 cm-1 range.

By studying the spectra at manyvalues of E

laser we were able to

show that the Raman spectra formetallic nanotubes are charac-teristically different from thoseof semiconducting nanotubes.This is shown in Fig. 4, wherethe broad spectra for1.55 <E

laser< 2.10 eV are identi-

fied with metallic nanotubes, andall the other spectra in the figureare identified with semiconduct-ing nanotubes.

At the present state of technol-ogy it is not possible to preparecarbon nanotubes samples thathave only a single nanotube di-ameter and a single orientation ofthe hexagons (chirality), nor canwe prepare a sample that has onlymetallic or semiconductingnanotubes. However, by carefulselection of the laser Ramanexcitation energy, we have beenable to carry out resonanceRaman experiments to study onlymetallic or semiconducting nan-otubes, even though the nanotubesample contains an abundance ofboth types. This distinction can

be made for two reasons. First, as shown inFig. 4., metallic and semiconducting nano-tubes have very different and distinct Ramanspectra. Second, large Raman intensities areobserved when the laser excitation energy isequal to the energy separation between sin-gularities in the 1-D electron density ofstates in the valence and conduction bands.Due to the remarkable electronic propertiesof the nanotubes, the density of states formetallic and semiconducting nanotubes arevery different, as shown in Fig. 3. If we nowplot all possible interband transitions up to3.0 eV for all possible nanotubes with diam-eters dt

in the range dt < 3.0 nm, we obtain amap (Fig. 5) showing the metallic window(shaded region) for nanotubes with a spec-ified diameter distribution. In resonantRaman scattering both the incident and scat-tered photons can be in resonance with aninterband transition.

Additional information on nanotubes isprovided by exploiting the anti-Stokes Ra-man peaks, which occur at the high-energyside of the excitation laser frequency. Ingeneral, according to simple Raman theoryand demonstrated by many experiments,one expects the anti-Stokes spectrum todisplay the same Raman shifts as the Stokesspectrum, which occurs at the low energy

continued on page 4

����� THE SPECTROGRAPH

Published by the George R. Harrison Spectros-copy Laboratory at the Massachusetts Institute ofTechnology, Cambridge, MA 02139-4307. Com-ments, suggestions, and inquiries can be directed tothe editor.

Editor: Doreen M. CharbonneauReporter: Alison M. Hearn

����� GEORGE R. HARRISONSPECTROSCOPY LABORATORY

Director: Michael S. Feld

Associate Director for Scientific Coordination:Jeffrey I. Steinfeld

Associate Director for Project Coordination:Ramachandra R. Dasari

The Spectroscopy Laboratory houses two laser researchresource facilities. The MIT Laser Research Facility, sup-ported by the National Science Foundation, provides sharedfacilities for core researchers to carry out basic laser researchin the physical sciences. The MIT Laser Biomedical Re-search Center, a National Institutes of Health BiomedicalResearch Technology Center, is a resource center for laserbiomedical studies. The LBRC supports core and collabora-tive research in technological research and development. Inaddition, it provides advanced laser instrumentation, alongwith technical and scientific support, free of charge touniversity, industrial and medical researchers for publish-able research projects. More information can be found on ourweb site //http:web.mit.edu/spectroscopy/www/. Write orcall for further information or to receive our mailings,(617) 253-4881.

Figure 3: 1-D electronic density of states calculated with a tight bindingmodel for (8,8), (9,9), (10,10), (11,11), and (12,12) armchair nanotubesand for (14,0), (15,0), (16,0), (17,0), and (18,0) zigzag nanotubes. [J.C.Charlier] Wave vector conserving optical transitions can occur betweenmirror image singularities in the 1-D electronic density of states of thevalence and conduction bands of the nanotubes, i.e., v

1 -› c

1 and v

2 -› c

2,

etc. These optical transitions, which are given in the figure in units ofeV and have the corresponding energies of E

11, E

22, etc., are responsible

for the resonance Raman effect.

Figure 4: First-order Raman spectra for the tangentialband of carbon nanotubes (1500-1650 cm-1), taken fornine laser excitation energies. At lower frequencies(1300-1350 cm-1) the D-band feature is observed, andat higher frequencies ~1750 cm-1 a combination modeis seen for some values of E

laser.

Page 3

Mildred Dresselhaus is one of the bestknown professors at MIT. She holds facultypositions in both the Department of Electri-cal Engineering and Computer Science andthe Physics Department, and she i an Insti-tute Professor, a special honor reserved foronly a few MIT faculty members who distin-guish themselves amongst their peers. Shewas President of the American Physical So-ciety and directed the Center for MaterialScience for a number of years. Millie, as sheis affectionately known throughout our com-munity, is also well-known for her veryeffective efforts in educating and advancingwomen and minority students and scientists.Her nanotube research in the SpectroscopyLaboratory is stimulating to all.

Millie was born and grew up in New YorkCity. She received her undergraduate educa-tion at Hunter College in New York City,and then spent a year of study at CambridgeUniversity and another year at Harvard Uni-versity. After a Ph.D. and postdoc at theUniversity of Chicago and Cornell Univer-sity in the area of microwave studies ofsuperconductivity, she arrived at MIT in1960 as a research staff member at the SolidState Division at the MIT Lincoln Laborato-ry. Her move to Lincoln Laboratorycoincided with her switching research fieldsto magneto-optic studies of semiconductorsand semimetals at high magnetic fields. In1962 she conducted the first magneto-opticstudies of graphite, which elucidated manynew aspects of the electronic structure of thisremarkable semimetal. To distinguish be-tween the spectra for left and right circularlypolarized radiation, higher resolution was

����� Personalities

Mildred S. Dresselhaus

needed. So, in 1966 she joined forces withProfessor Ali Javan to do the first high fieldmagneto-optic experiment with a lasersource, with Paul R. Schroeder as their joint-ly supervised graduate student. Paul’s Ph.D.thesis not only broke new ground in magne-to-optic technology, but led to a majoradvance in our understanding of the elec-tronic structure of graphite as we know ittoday, with the first correct identification ofthe location of electrons and holes in theBrillouin zone.

As a result of this collaboration, Milliebecame acquainted with many members ofProfessor Javan’s research group, includingMichael Feld and Ramachandra Dasari, wholater became directors of the SpectroscopyLaboratory. In her autobiographical mate-rial, Millie wrote that the SpectroscopyLaboratory has contributed greatly to theresearch capabilities of many investigatorsat MIT and elsewhere.

After joining the MIT faculty in 1967, firstas the Abby Rockefeller Mauze visiting pro-fessor of Electrical Engineering, and then in1968 as a permanent member of the Engi-neering School faculty, collaboration withthe Javan group continued for a few years.Research in the Dresselhaus group shifted tomany different areas of solid state physics,but including topics in carbon science andspectroscopy, such as studies of the struc-ture and properties of carbon fibers, graphiteintercalation compounds, liquid carbon acti-vated porous carbons, carbon aerogels,carbon fullerenes and carbon nanotubes. Thislong history in carbon-based materials hasestablished MIT as a world center in thisresearch area. Many collaborators and manyvisitors world-wide have come to work forperiods of time with the Dresselhaus groupin the area of carbon science. Over the years,Raman spectroscopy has become establishedas a major characterization tool for carbonbased materials, whether in the form of graph-ite, disordered carbons, porous carbons, andmore recently fullerenes and carbon nano-tubes. Millie writes that, “through ourassociation with the George R. HarrisonSpectroscopy Laboratory, it has been possi-ble to break new ground in exploring thedifferences between the spectra for metallicand semiconducting nanotubes, differencesin the Stokes and anti-Stokes spectra, iden-tification of the characteristics of the 1Delectron density of states, identification of

the mechanism behind the unusual proper-ties of the so-called zone edge D-band andits second-order G’ band counterpart in sp2

carbons and carbon nanotubes, and finally instudy of the enormous enhancements in theintensity of the Raman spectra of single wallcarbon nanotubes made possible by surface-enhanced Raman scattering using fractalcolloidal silversubstrates in collaborationwith Katrin Kneipp, Sandra D. M. Brown, P.Corio, M. A. Pimenta, A. Marucci and G.Dresselhaus.”

Millie joined the MIT Department of Phys-ics in 1983, and was named an InstituteProfessor at MIT in 1985. She is a memberof the American Philosophical Society, theNational Academy of Sciences, the NationalAcademy of Engineering, and a Fellow ofthe American Academy of Arts and Scienc-es, the American Physical Society, the IEEE,the Materials Research Society, the Societyof Women Engineers, and the AmericanAssociation for the Advancement of Sci-ence. She has served as President of theAmerican Physical Society, as Treasurer ofthe National Academy of Sciences, and aspresident of the American Association forthe Advancement of Science. She is onnumerous advisory committees and coun-cils, and is a foreign member of the JapaneseAcademy of Engineering. She has receivednumerous awards, including the KillianAward in 1986, the National Medal of Sci-ence from President George Bush in 1990,and 16 honorary doctorates. She is the co-author of three books on carbon science.

Millie is featured as one of the hundredmost important women in the United Statesin the November, 1999 issue of Ladies’Home Journal. In the article, her humanisticlove of science is apparent in her reasons forspeaking to her granddaughter’s third-gradeclass: “If children don’t get exposed to sci-ence, they will miss out on something that isso useful for living and for thinking abouthow things work.” Millie showed the chil-dren how a soccer ball is a real-life model ofa carbon fullerene.

Millie lives in Lexington with her hus-band (and research colleague) Gene, a seniorscientist working at MIT’s Magnet Labora-tory. They have four grown children. Forrelaxation, Millie is an enthusiastic chambermusic player, playing both violin and viola.

�

Page 4

Figure 5: Calculation of the energy separations (∆E = Eii(d

t) for all carbon nanotubes vs. diameters in the range

0.7 < dt < 3.0 nm and E

ii(d

t) < 3.0 eV. Semiconducting and metallic nanotubes are indicated by crosses and open

circles, respectively. The filled squares denote zigzag nanotubes. The dashed horizontal lines correspond to Elaser

values at which Raman spectra were taken, and the vertical lines denote dt = 1.49± 0.2 nm for one of our samples.

The inset is the calculated IM(Elaser

,d0) for the resonance window for metallic nanotubes for the Stokes (solid curve)

and anti-Stokes (dashed curve) scattering processes. The arrow in the inset at 1.69 eV (at the center of the metallicwindow) is drawn as a solid horizontal line in the main figure. The cross-hatch pattern is the E

laser range for which

the equation predicts that metallic nanotubes in both the Stokes and anti-Stokes spectra.

side of the excitation laser. However, forSWNT’s we observe Stokes/anti-Stokes fre-quency asymmetries. These arise from thefact that Stokes and anti Stokes scatteringselectively probe metallic and semiconduct-ing SWNT’s, due to different resonanceRaman scattering conditions for semicon-ducting and metallic nanotubes [2,3]. Toselect appropriate values of E

laser to access

only metallic nanotubes in a given sample,with diameter distribution determined frommeasurement of the radial breathing modefrequencies, we use the plots in the inset toFig. 5. For the range of E

laser common to both

the Stokes (S) and anti-Stokes (AS) curvesin the inset, we would expect both the Stokesand anti-Stokes spectra to exhibit contribu-tions from metallic nanotubes, while for thelower (higher) E

laser values within the metal-

lic window only the metallic nanotubescontribute to the anti-Stokes spectrum.

As a consequence, we show in Fig. 6 theStokes (phonon emission) and anti-Stokes(phonon absorption) Raman spectra for fourspecially selected laser excitation energiesfor a nanotube sample with a diameter dis-

Resonance...continued from page 2

tribution of 1.49 ± 0.20 nm. The figureshows that at E

laser = 2.19 eV only the semi-

conducting nanotubes are in resonance, whileat E

laser = 1.58 eV only the metallic nanotubes

are in resonance for both Stokes and anti-Stokes processes.

continued on page 5

In general, the Stokes and anti-StokesRaman spectra for molecules and crystal-line solids are the same. However, theunusual nanotube-selective resonant Ra-man process that is operative for carbonnanotubes can give rise to Stokes spectrathat are very different from the correspond-ing anti-Stokes spectra for the same valueof E

laser. This effect is shown in the spectra

of Fig. 6. Here, spectra are shown at Elaser

=1.92 eV, where the Stokes spectrum is formetallic nanotubes and the anti-Stokes spec-trum is for semiconducting nanotubes,whereas at E

laser = 1.49 eV, the opposite

situation occurs, with the semiconductingnanotubes in resonance for the Stokes pro-cess and the metallic nanotubes in resonancefor the anti-Stokes process. [2,3] A Lorent-zian lineshape analysis of the Stokes andanti-Stokes spectra at E

laser = 1.49 eV (Fig.

7) shows that the Stokes spectrum containsLorentzian components only for semicon-ducting nanotubes, whereas the anti-Stokesspectrum shows Lorentzian components as-sociated only with metallic nanotubes. Incontrast, at E

laser = 1.92 eV only semicon-

ducting nanotubes contribute to theanti-Stokes spectrum, whereas the Stokesspectrum contains contributions from bothmetallic and semiconducting nanotubes.

Finally, SWNT’s can show a strong sur-face-enhanced Raman effect when in contactwith metallic structures of nanometer sizedroughness [2]. The strong resonant enhance-ment of the Raman signals open up exciting

Figure 6: Stokes and anti-stokes Raman spectra for SWNT’s taken at 4 different values of Elaser

.

Page 5

Figure 7: Lorentzian fits to the Stokes and anti-Stokes tangential

opportunities for studying the Raman spec-trum of a small number and, perhaps, evensingle a single nanotube. In particular, itshould be possible to retrieve the intrinsicproperties which in the above experimentsare hidden by the inhomogeneous broaden-ing resulting from ensemble averaging. Byusing fractal colloidal silver particles [2] insurface-enhanced Raman scattering (SERS)experiments, similar to those reported in

Spectrograph Vol. 14 #1 forsingle molecules detected bySERS, we have been able toobserve Raman spectra withnarrow linewidths from a verysmall number of carbon nano-tubes (Fig. 8), free of theinhomogeneous line broaden-ing generally found in normalresonant Raman scattering(~ 25 cm-1). Our estimate forthe natural linewidth of singlewall nanotubes is ~10 cm-1.The enhancement of the Ra-man signal by many orders ofmagnitude (1012) through theSERS process, together withstrong spatial confinement ofthis enhancement, shouldeventually allow observationof the Raman spectrum for anindividual single carbonnanotube, ~1 nm in diameter.

References:

[1] S. Brown, P. Corio, A. Marucci, M.Pimenta, M. Dresselhaus, and G.Dresselhaus, “Second-Order ResonantRaman Spectra of Single-Walled CarbonNanotubes”, Phys. Rev. B 60, 1999 (In press).

[2] K. Kneipp, H. Kneipp, P. Corio, S. D. M.Brown, K. Shafer, J. Motz, L. Perelman, E.B. Hanlon, A. Marucci, G. Dresselhaus, and

M. Dresselhaus, “Surface-Enhanced and Normal Stokesand Anti-Stokes RamanSpectroscopy of Single-walled Carbon Nanotubes”,Phys. Rev. Lett., 1999 (Inpress).

[3] S. Brown, P. Corio, A.Marucci, M. Dresselhaus, M.Pimenta, and K. Kneipp,“Anti-Stokes Raman Spectraof Single-Walled CarbonNanotubes”, Phys. Rev. BRapid Comm., 1999 (Inpress). �

Alzheimer’s...continued from page 1

a basis for a new diagnostic technique.

The major advantage of NIR optical tech-niques is their potential to provide adefinitive diagnosis non-invasively, usingsafe, portable, low cost technology basedon laser spectroscopy. The human body istranslucent in the NIR region of the opticalspectrum, and NIR light penetrates the adulthead and can be used for non-invasivediagnostic purposes, such as cerebral oxy-gen monitoring. [1] The diagnostic signalsobserved with NIR fluorescence andRaman spectroscopy are biomolecular inorigin, and therefore carry specific infor-mation about the disease. This informationcan be used not only to diagnose AD, butalso to gain insight into the mechanismsand progression of the disease in vivo. Suchinsight would be of utmost utility in thedevelopment of pharmaceutical interven-tions for AD, to aid drug design, and tomonitor the effects of candidate drug treat-ments on living patients during trials.

As a first step, we investigated whetherbrain tissue from AD patients could bedifferentiated from that of age matchednon-AD controls in vitro. We have exam-ined both NIR fluorescence and NIR Ramanspectroscopy.

For the fluorescence study, unfixed tem-poral cortex specimens from AD cases andage-matched, non-AD controls were fro-zen at autopsy and then thawed just prior tospectral measurement. Age matched con-trols are especially important because thebrain tissue has undergone all the changesassociated with normal aging, such as dif-fuse amyloid deposition, arteriosclerosis,etc. - changes that must be distinguishablespectroscopically from those associatedwith AD if a practical diagnostic is to berealized. Spectra of intrinsic tissue fluores-cence induced by 647nm light were recordedfrom 650 to 850 nm (Fig. 1). We usedprincipal component analysis of the tissuespectra from 17 AD cases and 5 non-ADcontrol cases in a calibration study to estab-lish a diagnostic algorithm. The principalcomponents are a set of spectral “buildingblocks” which can be linearly combined toaccurately describe all of the tissue spectra,both normal and diseased. [2] Retrospec-tively applied to the calibration set, thealgorithm correctly classified 23 of 24 spec-

Figure 8: Typical fractal colloidal silver clusters (see inset) and selectedSERS spectra of the tangential band of single wall carbon nanotubes,with linewidths as small as 9.5 cm-1, collected from such a cluster area

Resonance...continued from page 4

continued on page 8

Page 6

Lester Wolfe Workshops in Laser Biomedicine

Future Directions of Optics in MedicineDiagnosis, Pathology, Imaging, Therapy

Friday, November 5, 1999, 4:00-7:00 PM

Massachusetts Institute of Technology, Room 6-12077 Massachusetts Avenue, Cambridge

Diagnosis: Mining Optical Information Brian C. Wilson, Ontario Cancer Institute

Spectral Diagnosis: Learning from Pathology’s Mistakes Maryann Fitzmaurice, University Hospitals of Cleveland

Hide and Seek in Medical Imaging Steven L. Jacques, Oregon Medical Laser Center

Therapy: Trouble and Opportunities for the Medical Photon R. Rox Anderson, Wellman Laboratories of Photomedicine

Panel Discussion* Michael S. Feld, Moderator, MIT

*Prizes will be awarded for the best audience participation!

Refreshments at 3:30 P.M.Sponsored by MIT Laser Biomedical Research Center,

MGH Wellman Laboratories, MIT Industrial Liaison Program &Harvard-MIT Division of Health Sciences and Technology

PLEASE POST

Page 7

Seminar onMODERN OPTICS AND SPECTROSCOPY

FALL SEMESTER 1999

September 21 David E. Pritchard, MITAccurate Atomic Masses for Metrology and Fundamental Physics

October 5 Tilman Pfau, MITAtom Optics

October 12 Roberto Onofrio, University of PadovaMeasuring Casimir and Gravitational Forces Using Micromechanical Resonators

October 19 Kevin Lehmann, Princeton UniversitySpectroscopy and Dynamics in Superfluid 4He Nanodroplets

October 26 Ernst Otten, University of MainzNeutrino Mass Measurement from T

2 Decay

November 2 George Flynn, Columbia UniversityOrder, Orientation and Chirality at the Liquid/Solid Interface;The View from a Few Billionths of an Inch

November 9 Peter So, MITModern Optical Microscopy: Catching Biology in Action

November 16 James Anderson, Harvard UniversityChemistry, Radiation and Climate: Laser Development for a New Approach

November 23 C. Bradley Moore, University of California at BerkeleyVibrational Spectra of Dissociating Molecules

November 30 David S. Hall, JILA/ University of Colorado at Boulder; Amherst CollegeIntertwined Bose Einstein Condensates

December 7 Deborah S. Jin, JILA/ University of Colorado at BoulderA Quantum Degenerate Gas of Fermionic Atoms

TUESDAYS, 12:00-1:00, Marlar Lounge (37-252), Ronald E. McNair BuildingFor map see <http://amo.mit.edu/mos.html>Refreshments served following the seminar

Sponsored by George R. Harrison Spectroscopy Laboratory andResearch Laboratory of Electronics,

Massachusetts Institute of Technology andRowland Institute for Science

PLEASE POST

Page 8

imens. We then applied this algorithm to aprospective set comprised of 19 specimensfrom 5 AD brains and 2 non-AD controlbrains, none of which were included thecalibration study. We found that three ofthe five control specimens and all AD spec-imens were correctly diagnosed.

The diagnostic algorithm [3] uses a lin-ear combination of the first two principalcomponent contributions, scores 1 and 2, toyield a numerical value, S, associated witheach specimen. Positive values of S indi-cate a positive diagnosis for AD based onthe specimen’s NIR fluorescence spectrum;negative S values indicate that no ADpathology has been detected spectroscopi-cally. In the prospective validation results,Fig. 2, open squares indicate that the spec-imen is from a brain withneuropathologically confirmed AD, andsolid squares indicate that the specimen isfrom the non-AD control group. Twenty-three of 24 calibration set specimens and17 of 19 validation set specimens were

continued on page 9

Figure 1. Brain tissue fluorescence.

Figure 2. Validation study results.

correctly diag-n o s e dspectroscopical-ly. [3]

In addition todifferentiatingdiseased fromcontrol tissue,NIR fluores-cence resultspoint to thepossibility of ex-tracting otherdetails of the tis-

sue pathology. Among the controlspecimens of the calibration set, thosewith notable diffuseamyloid depositionshowed the most neg-ative scores, and hada distinct fluores-cence lineshape. Theability to distinguishdiffuse amyloiddeposits from the fib-rillated amyloid ofsenile plaques may beessential for detect-ing the onset of AD,since amyloid depo-sition occursnormally with aging,even in the absenceof dementia. Thisanalysis also revealedinformation con-tained in the fluorescence spectraregarding the pathophysiological severi-

Figure 3. Brain tissue Raman spectrum.

ty of the disease. Pathology reports indicat-ed AD severity for nine calibration setspecimens. Of these, 6 of 7 “AD severe”specimens had S values above 0.04 and 2 of2 “AD moderate” specimens had values0.03>S>0.02. This result, though prelimi-nary, is an important indication of thepotential for monitoring disease progressionand the efficacy of treatment interventions.

In a separate study, [2] NIR Raman spec-tra (Fig. 3) were acquired using 5 specimensfrom one control brain and 12 specimensfrom four AD brains. None of the brainsused in the Raman study were used for thefluorescence study. Raman measurementswere made using 830 nm excitation. Princi-

pal component analysis of the normalized,mean-centered Raman spectra followed bylogistic regression analysis of the principalcomponent scores revealed that a modelbased on score 2 and score 6 could properlyclassify this limited data set with 100%accuracy (Fig. 4).

The next step in this research is to estab-lish that the diagnostic content of theobserved NIR spectroscopic signals frombrain tissue can be extracted from spectraacquired non-invasively, in situ. First, suffi-cient light must reach the brain to inducemeasurable spectroscopic signals. Second,brain tissue spectra that reach the detector,modulated by the absorption, scattering andfluorescence effects of the intervening tis-sue (skin, skull, dura, arachnoid and

Alzheimer’s...continued from page 5

Page 9

cerebrospinal fluid), must still be analyzablefor their diagnostic content. These pointshave been addressed in connection with non-invasive applications of NIR absorptionspectroscopy. [1] Estimates based on ourmeasurements of brain tissue spectra in vitroand the optical transport coefficients report-ed in the literature for adult head in vivoindicate that in the specific case of non-invasive NIR fluorescence measurementsfor AD diagnosis, sufficient light will reachthe brain to induce measurable tissue fluo-rescence that is analyzable for its diagnosticcontent.

To evaluate potential interference fromtemporal bone during in vivo measurements,we acquired NIR fluorescence spectra of afresh autopsy specimen under conditionsidentical to the brain tissue fluorescencemeasurements. The bone spectrum was add-ed to that of brain tissue, and these compositespectra were analyzed using the same diag-nostic algorithm used in the NIR fluorescencestudy. The result was that the algorithmcould still distinguish between AD tissueand non-AD control tissue, and each speci-men’s classification was consistent with itsearlier result, despite the contribution ofbone fluorescence to the observed spectra.

The brunt of pathological alterations inAD occurs in association cortex, in particu-lar the temporal lobe. The bony covering ofthe skull is at its thinnest over the temporallobe, having a thickness of several millime-ters at that point. It is a matter of good

fortune that the brain region of interest indiagnosing AD is where the skull is thinnestand where it is the most convenient to irradi-ate non-invasively. To make themeasurement in vivo, an optical fiber probewill deliver low energy laser light and col-lect brain tissue fluorescence. Appropriatesource/detector geometry in the probe tomaximize the contribution of the brain tissuefluorescence to the detected signal will beemployed. [4] For these in vivo measure-ments, the optical fiber probe will be mountedin a headpiece that will position the probe atthe temple and that can be worn withoutdiscomfort for the required signal acquisi-

Figure 4. Raman study results.

tion time (Fig.5).

This research utilizes optical spectroscop-ic techniques that carry direct, molecularlevel information about disease. Both theexcitation light used and the measured braintissue autofluorescence and Raman spectraare at near-infrared wavelengths that canpropagate through skull and overlyingtissue. The results represent the first stepstowards developing a clinical tool that haspotential to be applied to the non-invasivediagnosis of Alzheimer’s disease in livingpatients.

References:

[1] Hielscher, et al., “Determination ofblood oxygenation in the brain by time re-solved reflectance spectroscopy (I): influenceof the skin, skull, and meninges”, SPIE Proc:Biochemical Diagnostic Instrumentation2136, 4 (1994).

[2] Hanlon, et al., “Prospects for in vivoRaman spectroscopy”, Phys. Med. Bio. 1999(In Press).

[3] Hanlon, et al., “Near infrared fluores-cence spectroscopy detects Alzheimer’sdisease in vitro”, Photochem. Photobio 70,236 (1999).

[4] Sevick, et al., “Photon migration in amodel of the head measured using time- andfrequency- domain techniques: potentials ofspectroscopy and imaging”, SPIE Proc: TimeResolved Spectroscopy and Imaging ofTissues 1431, 84 (1991). �

Figure 5. Portable NR Fluorescence System.

Alzheimer’s...continued from page 8

Page 10

����� SPECTROSCOPY LABORATORY PUBLICATIONS

“An Introduction to Lasers” Itzkan I andBourgelais DBC, in Lasers in Cutaneousand Aesthetic Surgery , eds: K.A. Arndt,J.S. Dover, and S.M. Olbricht,(Lippincott-Raven publishers), p. 3(1997).

“Analytical Method of EstimatingChemometric Prediction Uncertainty”,Berger AJ, Feld MS. AppliedSpectroscopy, 51, 725 (1997).

“Autofluorescence Characterization of OralMucosa”, Ingrams DR, Dhingra JK, RoyK, Perrault DF, Bottrill ID, Kabani S,Rebeiz EE, Pankratov MM, ShapshaySM, Manoharan R, Itzkan I, Feld MS.Head & Neck, 19, 27 (1997).

“Convolution Picture of the BoundaryConditions in Photon Migration and ItsImplication in Time-Resolved OpticalImaging of Biological Tissues”, Wu J.Optical Society of America A featuredissue “Diffusion Photons in TurbidMedia”, 14, 280 (1997).

“Determination of Human Coronary ArteryComposition by Raman Spectroscopy”,Brennan JF, Romer TJ, Lees RS, TercyakAM, Kramer JR, Feld MS. Circulation,96, 99 (1997).

“Feasibility of Measuring Blood GlucoseConcentration by Near-Infrared RamanSpectroscopy”, Berger AJ, Itzkan I, FeldMS. Spectochimica Acta, 53: 2887(1997).

“Fluorescence Tomographic Imaging inTurbid Media Using Early ArrivingPhotons and Laplace Transforms” WuJ, Perelman L, Dasari RR, Feld MS.Proceedings of National Academic ofSciences, 94, 8783 (1997).

“History of Lasers in Medicine” Itzkan I andDrake EH, in Lasers in Cutaneous andAesthetic Surgery , eds: K.A. Arndt,J.S. Dover, and S.M. Olbricht,(Lippincott-Raven publishers), p. 11(1997).

“Investigation of Early Cancerous Changesin Bladder Tissue by Autofluorescence”,Arendt JT, Levin HS, Klein EA,Manoharan R, Feld MS, Cothren RM.Proceedings of the 19th AnnualInternational Conference of IEEE,Chicago, IL, 8.4.1-c, 2290 (1997).

“Near Infrared Raman Spectrometer Systemsfor Human Tissue Studies” Brennan JF,Wang Y, Dasari RR, Feld MS. AppliedSpectroscopy, 51, 201 (1997).

“Optical Bistability Induced by MirrorAbsorption: Measurement ofAbsorption Coefficients at the Sub-PPMLevel”, An K, Sones BA, Fang-Yen C,Dasari RR, and Feld MS. Optics Letters,22, 1433 (1997).

“Optical Tomographic Imaging UsingFluorescence Photon Migration”, ChenK, Perelman LT, Dasari RR, Feld MS.Proceedings of THICOLS ’97,Hangzhou, China, June 1-9, 421 (1997).

“Photon Migration of Near Diffusive Photonsin Turbid Media: A Lagrangian-BasedApproach”, Perelman LT, Winn J, WuJ, Dasari RR, Feld MS. J. Opt. Soc.Am., A 14, 224 (1997).

“Quantum Trajectory Analysis ofThresholdlike Transition in theMicrolaser” Yang C, An K. PhysicalReview, A 55, 4492 (1997).

“Semilcassical Four-Level Single-AtomLaser”, An K, Feld MS. PhysicalReview, A 56, 1662 (1997).

“Single Molecule Detection Using Surface-Enhanced Raman Scattering (SERS)”,Kneipp K, Wang Y, Kneipp H, PerelmanLV, Itzkan I, Dasari RR, Feld MS.Physical Review Letters, 78, 1667(1997).

“The Microlaser: A Quantized RabiOscillator” An K, Childs JJ, Yang C,Feld MS, Dasari RR. Proccedings ofthe International Conference onSpectroscopy: Perspectives andFrontiers, Bhabha Atomic ResearchCenter, ed: A.P. Roy, Bombay, India,p.70 (1997).

“Traveling-Wave Atom-Cavity Interactionin the Single Atom Microlaser”, An K,Dasari RR, Feld MS. Optic Letters, 22,1500 (1997).

“An Enhanced Algorithm for LinearMultivariate Calibration”, Berger AJ,Koo T-W, Itzkan I, Feld MS. AnalyticalChemistry, 70, 623 (1998).

“Biochemical Composition of HumanPeripheral Arteries Using Near InfraredRaman Spectroscopy”, Salenius JP,Brennan JF, Miller A, Wang Y, Aretz T,Sacks B, Dasari RR, Feld MS. Journalof Vascular Surgery, 27 (4) :710-719April, (1998).

“Detection and Identification of a singleDNA base molecule using surface-enhanced Raman scattering (SERS)”,Kneipp K, Kneipp H, Kartha VB,Manoharan R, Deinum G, Itzkan I,Dasari RR, Feld MS. Physical ReviewE, 57, 6281 (1998).

“Distribution of the Paths of Early-ArrivingPhotons Traversing a Turbid Medium”,Winn JN, Perelman LT, Chen K, Wu J,Dasari RR, Feld MS. Applied Optics,37, 8085 (1998)

“Early Diagnosis of Head & Neck SquamousCell Carcinoma in an Animal ModelUsing Fluorescence Spectroscopy”,Dhingra JK, Zhang X, McMillan K,Kabani S, Manoharan M, Itzkan I, FeldMS, Shapshay SM. Laryngoscope, 108,471 (1998).

Page 11

����� SPECTROSCOPY LABORATORY PUBLICATIONS

“Histopathology of Human CoronaryAtherosclerosis by Quantifying itsChemical Composition with RamanSpectroscopy”, Romer TJ, Brennan III,JF, Fitzmaurice M, Feldstein ML,Deinum G, Kramer JR, Lees RS, FeldMS. Circulation, 97, 878 (1998).

“Mathematical Model of FluorescenceEndoscopic Image Formation”, WangTD, Janes SG, Wang Y, Itzkan I, VanDam J, and Feld MS. Applied Optics,37, 8103 (1998).

“Measurement of Glucose in Human BloodSerum Using Raman Spectroscopy”,Koo T-W, Berger AJ, Itzkan I, HorowitzG, Feld MS. IEEE LEOS newsletter,12, 18 (1998).

“Near-Infrared Raman Spectroscopy ofHuman Whole Blood and Serum”,Berger AJ, Itzkan I, MS Feld.Proceedings of SPIE, Long Beach, CA,(1998).

“Near-Infrared Surface-Enhanced RamanScattering Can Detect Single Moleculesand Can Probe “Hot” VibrationalTransitions” Kneipp K, Kneipp H,Manoharan R, Itzkan I, Dasari RR, FeldMS. J. Raman Spectroscopy, 29, 743(1998).

“Observation of Periodic Fine Structure inReflectance from Biological Tissue: ANew Technique for Measuring NuclearSize Distribution”, Perelman LT,Backman V, Wallace M. Zonios G.Manoharan R, Nusrat A. Shields S, SeilerM, Lima C, Hamano T, Itzkan I, VanDam J, Crawford JM, MS Feld. Phys.Rev. Letters, 80, 627 (1998).

“Optical heterodyne detection of laser-induced gratings”, Maznev AA, NelsonKA, Rogers JA, Optics Letters, 23,1319 (1998)

“Raman Spectroscopy and FluorescencePhoton Migration for Breast CancerDiagnosis and Imaging” Manoharan R,Shafer K, Perelman L, Wu J, Chen K,Deinum G, Fitzmaurice M, Myles J,Crowe J, Dasari RR, Feld MS. Specialissue “Symposium in Print” ofPhotochemistry and Photobiology,“Innovations in Optical Techniques forBiodiagnostics”, 67, 15 (1998).

“Single Molecule Detection of a CyanineDye in Silver Colloidal Solution UsingNear-Infrared Surface-EnhancedRaman Scattering” Kneipp K, KneippH, Deinum G, Itzkan I, Dasari RR, FeldMS. Applied Spectroscopy, 52, 175(1998).

“Spectral Pathology”, Zonios G, CothrenRM, Crawford JM, Fitzmaurice M,Manoharan R, Van Dam J, Feld MS.Annals of the New York Academy ofSciences, 838, 108 (1998).

“Spectroscopic Characterization of a DNA-Binding Domain, Z alpha, from theEditing Enzyme, dsRNA AdenosineDeaminase: Evidence for Left-HandedZ-DNA in the Z alpha-DNA Complex”,Berger I, Winston W, Manoharan R,Schwartz T, Alfken J, Kim YG,Lowenhaupt K, Herbert A, Rich A.Biochemistry, 37, 13313 (1998).

“Surface-Enhanced Raman Scattering(SERS) - a New Tool for SingleMolecule Detection and Identification”Kneipp K, Kneipp H, Manoharan R,Itzkan I, Dasari RR, Feld MS.Bioimaging, 6, 104 (1998).

“The Single Atom Laser”, Feld MS, An K.Scientific American, 276, 56 (1998).

“Tomographic Detection of FluorophoresEmbedded in Tissue-Like Phantom”,Wu J, Perelman L, Dasari RR, and FeldMS. In the proceedings of theconference on Advances in OpticalImaging and Photon Migration,Orlando, Fl, March, in press, (1998).

“Analysis of Nucleotides and AromaticAmino Acids in Normal and NeoplasticColon Mucosa by Ultraviolet Resonanceand Raman Spectroscopy” Boustany N,Manoharan R, Feld MS, Dasari RR.Laboratory Investigation, in press(1999).

“Clinical Pathology”, Kramer JR, and FeldMS, in McGraw-Hill 1999 Yearbook ofScience and Technology, p. 79 (1999).

“Diffuse Reflectance Spectroscopy ofHuman Adenomatous Colon Polyps InVivo”, Zonios G, Perelman LT,Backman V, Manoharan R, FitzmauriceMA, Van Dam J, Feld MS. AppliedOptics, in press (1999).

“Histological Classification of RamanSpectra of Human CoronaryAtherosclerosis Using PrincipalComponent Analysis”, Deinum G,Rodriguez D. Romer TJ, FitzmauriceM, Kramer JR, Feld MS. AppliedSpectroscopy, 53, 938 (1999).

“In Vivo Identification of Colonic DysplasiaUsing Fluorescence EndoscopicImaging”, Wang TD, Crawford JM, FeldMS, Wang Y, Itzkan I, Van Dam J.Gastrointestinal Endoscopy, 49, 447(1999).

“Light Scattering for Biomedical Analysisand Disease Diagnosis,” Feld MS, inSpectroscopy of Biological Molecules:New Directions (8th EuropeanConference on the Spectroscopy ofBiological Molecules) J. Greve, G.J.Puppels, and C. Otto eds. (KluwerAcademic Publishers, Dordrecht-Boston-London), Fall 1999.

“Multicomponent Blood Analysis by Near-Infrared Raman Spectroscopy”, BergerAJ, Koo T-W, Itzakn I, and Feld MS.Applied Optics, in press (1999).

Page 12

����� SPECTROSCOPY LABORATORY PUBLICATIONS 1997-99

“Polarized Light Scattering Spectroscopyfor Quantitative Measurement ofEpithelial Cellular Structures In Situ”,Backman V, Gurjar R, Badizadegan K,Dasari RR, Itzkan I, Perelman LT, andFeld MS. IEEE, Journal of SelectedTopics in Quantum Electronics on Lasersin Medicine and Biology, 5, 1019(1999).

“Quantitative Analysis of Mucosal Tissuesin Patients Using Light ScatteringSpectroscopy”, Perelman LT, ZoniosG, Backman V, Gurjar R, Itzkan I, DasariRR, Van Dam J, Feld MS, in OpticalTomography and Spectroscopy of TissueIII, RR Alfano, B Chance, BJ Tromberg,eds. SPIE Press, 3597 (1999).

“Reagentless Blood Analysis by Near-Infrared Raman Spectroscopy” Koo T-W, Berger AJ, Itzkan I, Horowitz G,Feld MS. Diabetes Technology andTherapeutics, 1, 153 (1999).

“The Role of Laser-Induced FluorescenceSpectroscopy in the Detection of HumanAtherosclerosis”, van de Poll SWE,Dasari RR, Kramer JR. Current Science,77, 101 (1999).

“Spatial Coherenece of Forward ScatteredPhotons in a Turbid Medium” Yang C,An K, Perelman LT, Dasari RR, FeldMS. Journal of Optical Society ofAmerica, A 16, 866 (1999).

“Surface-Enhanced Non-Linear RamanScattering at the Single Molecule Level”,Kneipp K, Kneipp H, Itzkan I, DasariRR, Feld MS. Chemical Physics, inpress (1999).

“Resonant Raman Study ofPolyparaphenylene-based Carbons”Marucci A, Brown SDM, Pimenta MA,Matthews MJ, Dresselhaus MS,Nishimura K, Endo M. J. Mater. Res.,14, (1999).

“Study of the Overtones and CombinationBands in the Raman Spectra ofPolyparaphenylene (PPP)-basedCarbons” Marucci A, Pimenta MA,Brown SDM, Matthew MJ, DresselhausMS, Endo M., J. Mater. Res., 14, 3447(1999).

“Anti-Stokes Raman Spectra of Single-Walled Carbon Nanotubes” BrownSDM, Corio P, Marucci A, DresselhausMS, Pimenta MA, Kneipp K., Submittedto PRB Rapid Comm. (1999). �