Goldberg/Ünlü Group
Projects
- Rotating Aperature Interference Nanoscopy
- Spectral Self-Interference Fluoresence
Spectroscopy (conference paper)
- Optical Waveguide Phase Tomography
- Evanescent Waveguide Biosensing (project
page)
- High-Resolution Diffraction Grating
Fabrication
Independent Research
Pile
Shuffling
I've become interested in the mathematics of pile shuffling, which is
a technique for shuffling a deck of cards. One takes the top m cards
from a deck of n cards and so makes m separate piles. One then sequentially
places the next m cards from the n-m cards remaining in the deck on
top of the m cards, thus ending up with m piles of n/m cards, assuming
that m is a factor of n, or one ends up with m - n mod m piles with
one fewer card. The separate piles are then recombined into a deck of
n cards in a new ordering and the process is repeated. The technique
poses a number of interesting mathematical questions, since unlike the
riffle shuffle, even in practice this technique is completely deterministic.
How many shuffles are required to get a random ordering? How does the
size of n and m effect the result after multiple iterations? Can the
original ordering of the deck ever be regained by accident? Mathematica
is used heavily in this investigation.
2-D
Constant Aspect Ratio Strip Packing
The one dimensional bin packing problem is well researched. The goal
is to "pack" a set of one dimensional objects of varying lengths,
li, into the fewest bins each of a finite size L, such that all li <= L.
A number of variations on this problem exist, and it has been well studied
because of its important applications, particularly in the allocation
of computer memory. This problem can also be easily generalized to two
and higher dimensions. A related problem is that of two dimensional
strip packing, also sometimes known as the cutting stock problem because
of its application to clothing manufacturers who wish to minimize the
length of cloth they must use to cut out a variety of patterns. For
the case where the patterns are rectangular and the cloth is a finite,
given width the problem is well studied, and there exist a number of
excellent algorithms that give an upper bound on the length required
to fit all of the patterns with the least wasted space. I've become
interested in a novel variant of this problem that I call the constant
aspect ratio strip packing problem, wherein the rectangular patterns
are allowed to scale both larger and smaller, so long as the aspect
ratio is maintained. The goal is to pack a set of rectangles into an
area of a given aspect ratio such that the entire area is covered, and
so that there's a minimum variance in the percent size increase or decrease
of each of the component rectangles. This problem has not been studied
because it does not arise naturally in the study if computer architecture
nor does it have immediate industrial applications. However, it is an
inherently interesting problem, and does have some applications, such
as in the generation of a collage from a set of computer images, which
is how the problem was first suggested to me.
History
of the Electron
After a lecture demonstration in chemistry class, Professor John Straub
and I became interested in the history of the discovery of the electron,
and why certain experiments were not performed at the turn of the century.
We have been actively pursuing this topic over the last several years
and intend to publish the results of our research within two years.
Avogadro
Constant
In the course of our investigation concerning the development of the
electron Professor Straub and I became very interested various methods
of measuring Avogadro's Number (Loschmidt's Number is equivalent).
The enormous variety in both experimental and theoretical methods for
the calculation of this very important chemical constant are quite astonishing,
and also very illustrative of the history of chemistry in the late nineteenth
and early twentieth centuries. We are currently compiling a graph of
various measures of this constant over time, complete with the appropriate
error bars, to demonstrate the specifically non-monotonically converging
nature of chemical metrology.
Dissociation
of Sodium Chloride
My interest is the dissociation or solvation of sodium chloride has
grown out of a paper (Pavel Jungwirth, "How Many Waters does it
Take to Dissolve a Rock Salt Molecule?" Journal of Physical
Chemistry A, 104, 145, 2000) I read in the fall of 2000, as I was
taking a physical chemistry course. I investigated this problem both
by delving back into the experimental and theoretical literature, as
well as performing ab initio calculations of the gas-phase dissociation
in various water clusters, using PC-Spartan Plus, a software
package made by Wavefunction Inc. In this investigation I found that
Spartan is essentially useless for serious physical chemistry research,
although it does have very intriguing possibilities in the teaching
of organic, and possibly physical, chemistry. I also found that there
were many serious methodological errors in Jungwirth's paper, and that
the generally accepted criteria for dissociation of ions in solutions
is somewhat unsatisfactory. I've recently revisited this problem as
part of the graduate class on quantum mechanics I took in the fall of
2001, and I am performing new ab initio calculations with the far more
powerful Gaussian 98W, and with a better understanding of quantum
mechanics and electronic structure. The solvation of sodium chloride
is an example of a little-researched area of chemical physics that I
am very interested in, called mesoscopic chemistry, or the chemistry
of clusters. While behaviors and interaction at the macroscopic and
microscopic levels are comparatively well understood, this understanding
has not translated cleanly into intermediate regimes. For example, while
one can classify a macroscopic collection as belonging to a particular
phase of matter (solid, liquid, gas, etc.) and it is clear that a single
water molecule, for instance, cannot be so classified, it is unclear
what meaning this concept has for a water cluster. I believe that this
will be an increasingly important area of research.
Digital
Libraries
As a member of the Faculty of Arts and Sciences Standing Committee on
the Library at Harvard University, and an avid user of libraries, I've
thought a great deal about how libraries are changing with the advent
and implementation of new technologies, particularly digital imaging
and searching algorithms. My interest in "digital libraries"
is actually representative of a greater interest in computing in the
humanities in general. I've done some working performing frequency analyses
of Ancient Greek literature, comparing technical aspects of Plato's
imitations of Lysias' oratory in the Phaedrus, to Lysias' own corpus,
using Plato's other works as a control, to see if Plato was aiming for
fidelity or humor in his imitation of Lysias. I've also worked a bit
with Professor Mark Schiefsky of the Classics department at Harvard
University on his Archimedes project, and I am a contributor to the
Voice of the Shuttle internet project. I'm also interested in strategies
of searching the internet, and the place of the internet in contemporary
research.
College
Mathematics Education
While a student of calculus and differential equations some years ago,
I became interested in college mathematics education, specifically introductory
mathematics education. I wondered whether the standard two-semester
introductory sequence of single-variable differential and integral calculus
was still the best choice for the first college mathematics courses,
which not only asks questions about calculus and discrete mathematics
(the latter being a candidate to replace calculus as the standard introductory
course) but also of the aims of college mathematics educations for various
kinds of students. I am kept current in this area by the seminars on
mathematics education organized by Professor Daniel Goroff of the Derek
Bok Center for Teaching and Learning at Harvard University, and I am
in the process of developing a complete argument covering how discrete
mathematics better fulfills the purpose of college mathematics for many
students.
College
Chemistry Education
My interest in science education at the university level is related
to my interest in mathematics education in so far as I am trying to
rethink the purpose of science training for various kinds of university
students. But the particular concerns of chemistry education are quite
different from those of mathematics. I am interested in several accepts
of college chemistry education: the structure and variety of general
chemistry courses; the use of technology, specifically quantum mechanical
modeling software, in the teaching of introductory organic chemistry;
the place of and structure of physical chemistry in the curriculum;
the discontinuity between introductory and advanced textbooks.
Raphael
Semmes
An extremely well-read and well-educated Confederate solider, Semmes
was both an admiral and general for the CSA. His memoirs of both the
Mexican War and the Civil War are fascinating, and he holds an unusual
position as a modern solider and as a southern gentleman imbued with
notions of honour. I've written extensively, but as of yet unpublished,
concerning this most interesting figure from the American Civil War.
George
Johnstone Stoney
While researching the history of the discovery of the electron, I found
that an Irish scientist, G.J. Stoney had actually coined the term electron,
some years before J.J. Thomson's discovery of the particle. Stoney had
a fascinating career as an extremely creative scientist, but is often
not so well known any longer, and there's lots of biological research
yet to be done on his career.
Randomness
of Run Support
It is often said that the number of wins and losses that a baseball
pitcher accumulates over a year is a poor measure of that pitchers worth,
since the number of runs that his own team scores behind him is out
of his control. Sometimes, the number of runs that the opposing team
scores is also out of his control because of errors committed by his
defense. Thus the earned run average (ERA, the number of earned runs
the pitcher would allow in an average nine-inning game) is widely touted
as a more fair metric to evaluate the performance of a pitcher. But
some say that pitchers with a low ERA, few wins, and many losses, are
not just saddled with bad luck, but just don't quite know how to win--how
to turn up the intensity when the going gets tough. Some are more charitable
and suggest that teams play different behind certain pitchers--either
not scoring as many runs as they might otherwise because they know they
don't necessarily have to in order to win, or playing sloppy defense
because the pitcher takes an inordinately long time in between pitches.
I'm currently looking at a large set of statistical data to see how
random run support really is.
Seat
Location in Lectures
Do students learn better if they sit closer to the front of the lecture
hall? Will they learn better if forced to sit closer to the front? How
do final grades correlate with the number of absences in a class?
Historiography
of Greek Mathematics
What is the best way to segment and categorize ancient Greek mathematics?
Should practical arithmetic, geometry, number theory, logic, and geometric
algebraic be looked at as separate disciplines? Did the last even really
exist in Greek mathematics? This is at once the central question and
the starting question in the history of Greek mathematics, and one that
occupies a great deal of my thoughts, and a great deal of my academic
courses in the history of science. Ultimately, it is from the historiography
of Greek mathematics, that my "continuity theory" of science
originates.
Linguistics
of Computer Languages
Computer programming languages have a number of striking similarities
with natural languages in both form and development. However, unlike
natural languages where the focus is almost always on verbal communication,
programming languages, and most other formal languages (such as mathematical
notation, or symbolic logic) lack a well defined phonetic system and
require a completely unforgiving orthography. The study of programming
linguistics is interesting from a purely theoretical computer science
or theoretical linguistics standpoint, but also has important practical
application, as it becomes significantly more important for a greater
proportion of the population to learn and use computer programing languages.
Chinese
Expatriate Names
I've become interested in what kinds of personal names Chinese people
living outside of China use. I've made an ethnographic study of this
question, drawing on responses from a set of college-aged students of
Chinese ethnicity living in the greater Boston area. I am planning to
significantly expand the ethnographic aspect of this study, and combine
it with the significant theoretical apparatus that I've constructed
as a project in a Chinese anthropology course I took in the fall of
2000.
Byron
and Ancient Greece
A very long time ago, when I was in high school, I wrote a paper about
the influence of Ancient Greek literature and thought on Lord George
Gordon Noël Byron, the great English poet (and my favourite English
poet). I continue to be interested in this topic, although I am not
currently actively revising my paper.