Annals of Biomedical Engineering, Vol. 34, No.
1, January 2006 (

C 2006) pp. 314
DOI: 10.1007/s10439-005-9000-9
Engineering Luminescent Quantum Dots for In Vivo Molecular
and Cellular Imaging
ANDREW M. SMITH, GANG RUAN, MATTHEW N. RHYNER, and SHUMING NIE
Departments of Biomedical Engineering and Chemistry, Emory University and Georgia Institute of Technology, Atlanta, Georgia
(Received 16 January 2005; accepted 30 June 2005; published online: 1 February 2006)
AbstractSemiconductor quantum dots are luminescent
nanoparticles that are under intensive development for use as a
new class of optical imaging contrast agents. Their novel proper-
ties such as optical tunability, improved photostability, and mul-
ticolor light emission have opened new opportunities for imaging
living cells and in vivo animal models at unprecedented sensitivity
and spatial resolution. Combined with biomolecular engineering
strategies for tailoring the particle surfaces at the molecular level,
bioconjugated quantum dot probes are well suited for imaging
single-molecule dynamics in living cells, for monitoring protein
protein interactions within specific intracellular locations, and for
detecting diseased sites and organs in deep tissue. In this article,
we describe the engineering principles for preparing high-quality
quantum dots and for conjugating the dots to biomolecular ligands.
We also discuss recent advances in using quantum dots for in vivo
molecular and cellular imaging.
KeywordsNanoparticles, Nanotechnology, Fluorescence, Liv-
ing cells, Living animals, Molecular imaging, Cytotoxicity,
Cationic peptides, Bioconjugation, Dynamic light scattering.
ABBREVIATIONS
FRET fluorescence resonance energy transfer
PEG polyethylene glycol
QD quantum dot
RES reticuloendothelial system
TOP trioctylphosphine
TOPO trioctylphosphine oxide
INTRODUCTION
Scientists and engineers have developed devices and
technologies spanning a wide range of scales, from tow-
ering skyscrapers to microfabricated electronic circuits. A
notable gap, however, exists between micro-scale engineer-
ing and molecular-scale chemistry. This gap occurs in a size
range from a few nanometers to hundreds of nanometers,
Address correspondence to Shuming Nie , Departments of Biomedi-
cal Engineering and Chemistry, Emory University and Georgia Institute
of Technology, 1639 Pierce Drive, Suite 2001, Atlanta, Georgia 30322.
Electronic mail: snie@emory.edu
3
and is currently under intense study across many disci-
plines. For examples, electronics engineers aspire to fab-
ricate chips with nanometer-scale parts, physicists hope to
investigate the intersection between classical physics and
quantum mechanics, and chemists aim to study the inter-
actions of colloidal macromolecules. This broad interest
arises from the fact that many materials on the nanometer
length scale have unique properties that are not available
from either discrete molecules or bulk solids. This fascinat-
ing length scale has also been exploited by nature, because
proteins, nucleic acids, and other components of cells all
have nanoscale dimensions. It is thus not surprising that
the life sciences have been one of the first fields to ben-
efit from nanoscience and nanotechnology. For example,
semiconductor quantum dots (QDs) have been developed
as fluorescent labels for bioimaging applications. In this
review article, we examine the engineering principles of
quantum dot synthesis and the use of quantum dot probes
in biomolecular imaging at a variety of length scales, from
the nanometer to micrometer resolution of living cells up
to the centimeter to meter scale of entire organisms.
Quantum dots are nearly spherical semiconductor par-
ticles with diameters of the order of 210 nanometers, or
roughly 20010,000 atoms. Their semiconducting nature
and their size-confinement properties have made them very
attractive for use in optoelectronic devices, biological detec-
tion, and also as fundamental prototypes for the study of col-
loids and the size-dependent properties of nanomaterials.
Bulk semiconductors are characterized by a composition-
dependent bandgap energy, which is the minimum energy
required to excite an electron to an energy level above its
ground state, commonly through the absorption of a photon
of energy greater than the bandgap energy. Relaxation of
the excited electron back to its ground state may be accom-
panied by photon emission. Because the bandgap energy is
dependent on the particle size, the optical characteristics of
a QD can be tuned by adjusting its size.43 Figure 1 shows
the optical properties of CdSe QDs at four different sizes
(2.2, 2.9, 44.1, and 7.3 nm in diameter).
In comparison with organic dyes and fluorescent pro-
teins, QDs are about 20 times brighter, mainly due to their
0090-6964/06/0100-0003/0

C 2006 Biomedical Engineering Society
4 SMITH et al.
FIGURE 1. Size-dependent optical properties of cadmium se-
lenide QDs dispersed in chloroform, illustrating quantum con-
finement and size tunable fluorescence emission. (A) Fluores-
cence image of four vials of monodisperse QDs with sizes
ranging from 2.2 to 7.3 nm in diameter. This image was ob-
tained with ultraviolet lamp illumination at 365 nm. (B) Flu-
orescence spectra of the same four QD samples, excited at
400 nm. Narrow emission bands (2326 nm FWHM or full-width
at half-maximum) indicate narrow particle size distributions.
(C) Absorption spectra of the same four QD samples. Notice
that the onset of absorption is slightly blue-shifted from the
emission peak for each QD sample, and that the absorption
spectra are very broad, so that a wide spectrum can be used
for excitation. Both the absorption and emission intensities
are plotted in arbitrary units (AU).
large absorption cross sections, 1001000 times more stable
against photobleaching, and show narrower and more sym-
metric emission spectra. In addition, a single light source
can be used to excite multicolor QDs, and the emission
wavelength can be tuned from the ultraviolet,64 throughout
the visible and near-infrared spectra,5,28,32,52 and even into
the mid-infrared.50 On the other hand, QDs are macro-
molecules that are an order of magnitude larger than or-
ganic dyes. For this reason, small dyes will be more ap-
propriate biological labels when the size of the fluorescent
label must be minimized. However, the similarity in size
between QDs and proteins also makes them biologically
relevant, and one of the first biological applications was the
use of QDs for protein detection.38
QD SYNTHESIS AND ENGINEERING
QD synthesis was first described in 1982 by Efros and
Ekimov,17,18 who grew nanocrystals and microcrystals of
semiconductors in glass matrices. Since their work, a wide
variety of synthetic methods for QDs have been described,
including preparation in aqueous solution at room tempera-
ture, synthesis at elevated temperature and pressure in an au-
toclave, and vapor-phase deposition on a solid substrate.4,11
Most syntheses yielding colloidal suspensions of QDs in-
volve the introduction of semiconductor precursors under
conditions that thermodynamically favor crystal growth, in
the presence of semiconductor-binding agents, which func-
tion to kinetically control crystal growth to maintain their
size within the nanoscale.
Because the size-dependent properties of QDs are most
pronounced when the nanoparticles are monodispersed,
there has been a drive to produce QDs with narrow size
distributions. Major progress toward this goal was made
in 1993 by Bawendis group, with the introduction of a
synthetic method for monodisperse QDs (<5% root-mean-
square in diameter) made from cadmium sulfide (CdS),
cadmium selenide (CdSe), or cadmium telluride (CdTe).42
Since this seminal report, the synthesis of CdSe QDs has
progressed rapidly, generating QDs that can span the visible
spectrum. As a result, CdSe has become the most common
chemical composition for QD synthesis, especially for bi-
ological applications. Many techniques have been imple-
mented to post-synthetically modify QDs for various pur-
poses, such as coating with a protective inorganic shell,12,27
surface modification to render colloidal stability,23,25 and
direct linkage to biologically active molecules.8,9 QD pro-
duction has now progressed from a chemical science to an
elaborate molecular engineering process. The most com-
mon scheme involves four steps: (1) synthesis of the QD
core, most often CdSe, in a high-temperature organic sol-
vent; (2) growth of an inorganic shell (usually zinc sulfide,
ZnS) epitaxially on the core to protect the optical properties
of the QD; (3) phase transfer of the QD from organic liquid
phase to aqueous solution; and (4) linkage of biologically
active molecules to the QD surface to render functional-
ity, or linkage of biologically inert polymers to minimize
biological activity. Each of these four steps is discussed
later.
Core Synthesis
The most effective and reproducible synthesis pro-
cedure for monodisperse QDs involves the addition of
semiconductor precursors to a liquid coordinating solvent at
high temperature, first described by Murray et al.42 The co-
 Engineering Luminescent Quantum Dots for In Vivo Molecular and Cellular Imaging
ordinating solvent commonly consists of trioctylphosphine
oxide (TOPO) and trioctylphosphine (TOP), which contain
basic functional groups that can bond to the QD surface
during growth to prevent the formation of bulk semicon-
ductors. The alkyl chains from coordinating ligands extend
away from the QD surface, rendering the QDs sterically
stable as colloids, dispersible in many nonpolar solvents. In
a typical synthesis of CdSe, room-temperature QD precur-
sors, dimethylcadmium and elemental selenium dissolved
in liquid TOP, are swiftly injected into hot (290350 C)
TOPO, immediately initiating nucleation of QD crystals.
CdSe nucleation and growth is favored thermodynamically,
because the precursors are introduced at concentrations well
above the solubility of the resulting semiconductor. How-
ever, crystal growth is kinetically controlled by monomer
diffusion, due to the high viscosity of the solvent (14 cp at
the melting point of TOPO, 50 C), and is also controlled
through the reaction rate of monomers at the QD surface,
due to strong binding of the coordinating solvent with semi-
conductor precursors and QD surfaces. A high temperature
at injection overcomes the steric, kinetic barrier, allow-
ing precursor association and nucleation. The swift drop in
temperature, combined with the drop in monomer concen-
tration (due to the nucleation of many small QD crystals),
stops nucleation within seconds after injection, allowing
even and homogeneous growth on similarly sized nuclei.
This separation of nucleation and growth is responsible for
the monodispersity of the final QDs. Also the use of a hot
solvent yields semiconductor nanoparticles that are highly
crystalline, while minimizing thermodynamically unfavor-
able lattice defects.
The initial step of nucleation is so rapid that it is difficult
to study experimentally, but the ensuing growth process
is controllable and proceeds slowly, and distinct stages of
growth have been identified. After nucleation, remaining
monomers grow epitaxially on the nuclei, and the QDs
reach a size-focusing point, at which the size distribution is
the narrowest.49 Interestingly, the quantum yield of the QDs
also reaches a high value, at which the QDs in the reaction
reach a bright point, and afterward decrease in photo-
luminescence efficiency.52 After this size-focusing point,
the monomers become depleted, and the size distribution
widens (defocusing). A process known as Ostwald ripening
also starts to occur, in which smaller particles are dissolved
to grow on large particles.49 These growth phases are depen-
dent on the monomer concentration in solution, and many
parameters (temperature, each precursors initial concen-
tration, and solvent composition) can be tuned to adjust the
size at which the QDs are focused. At the focusing point, it is
desirable to quench the reaction, usually by decreasing the
temperature until crystal growth is negligible. This synthe-
sis procedure has been found to be the most versatile and re-
producible for QDs composed of CdSe, although QDs with
other compositions have similarly been synthesized in co-
ordinating solvents.5,58,65 New variations on this synthesis
5
have used diverse molecular precursors for CdSe (cadmium
oxide, dimethylcadmium, and cadmium acetate, combined
with TOP-Se),42,48,53 various coordinating ligands (alky-
lamines and alkanoic acids) for enhanced monodispersity
and quantum yield,53,57 and it has been found that the co-
ordinating solvent may be replaced with a noncoordinating
solvent, like octadecene, containing a small amount of coor-
dinating ligand.62 The vast number of potential parameter
combinations has allowed the simple synthesis of CdSe
QDs with diameters between 2 and 8 nm, with the emission
wavelengths (450650 nm) spanning the entire visible spec-
trum with just one composition. By also adjusting the QD
composition (ZnS, CdS, CdSe, CdTe, PbS, PbSe, and their
alloys), it is now possible to span the wavelength range
4004000 nm.5,28,32,50,52,64 Adjusting the solvent charac-
teristics and initial precursor concentration has further re-
sulted in nanocrystals with diverse shapes, like rods and
tetrapods.47
When designing QD cores for a specific wavelength re-
gion, the first choice is to select a chemical composition,
since high-quality QDs can only be obtained for a certain
wavelength range for each composition. For example, CdSe
QDs may be tuned to emit between 450 and 650 nm, while
CdTe QDs may be tuned to emit from 500 to 750 nm.
The QD diameter is then chosen to determine the specific
wavelength of emission, and the QDs are then generated
through focused particle growth using the synthesis pa-
rameters described earlier. The resulting QDs are coated
in coordinating ligands and suspended in a crude mix-
ture of the coordinating solvent and molecular precursors.
The QDs are highly hydrophobic, and can be isolated and
purified from the reaction mixture, either through liquid
liquid extraction (a mixture of hexane and methanol),63 or
through precipitation from a polar solvent (methanol or ace-
tone) that dissolves the reactants and coordinating ligands,
but not the QDs.42 The pure core QDs are then used as
substrates for further modification to generate biological
probes.
Shell Growth
Because QDs have high surface area to volume ratios,
a large fraction of the constituent atoms are exposed to the
surface, and therefore have atomic or molecular orbitals
that are not completely bonded. These dangling orbitals
may form bonds with organic ligands such as TOPO. This
leads to an electrically insulating monolayer that serves
to passivate the QD surface by maintaining the inter-
nal lattice structure and protecting the inorganic surface
from external effects.3 However, the bond strength between
the organic ligand and the semiconductor surface atom is
typically much lower than the internal bond strength of
the semiconductor lattice, and desorption of ligands makes
the core physically accessible. For this reason, it is ad-
vantageous to grow a shell of another semiconductor on
6 SMITH et al.
the QD surface after synthesis. By using a shell of wider
bandgap than the underlying core, strong electronic insu-
lation results in enhanced photoluminescence efficiency,
and a stable shell provides a physical barrier to degrada-
tion or oxidation. As an example, to passivate CdSe QDs
with ZnS, the cores are purified to remove unreacted cad-
mium or selenium precursors, and then resuspended in a
coordinating solvent.12 Molecular precursors of the shell,
usually diethylzinc and hexamethyldisilathiane dissolved
in TOP, are then slowly added at elevated temperatures.
The temperature for growth of ZnS on CdSe is chosen such
that it is high enough to favor epitaxial crystalline growth,
but low enough to prevent nucleation of ZnS crystals and
Ostwald ripening of CdSe cores. Normally, this is a tem-
perature around 160220 C. The (core)shell (CdSe)ZnS
nanocrystals may then be purified just like the cores. Al-
though shell growth is a common procedure, uncapped
CdSe cores are also of great interest, especially for fluo-
rescence resonance energy transfer (FRET) applications,
in which physical access to charge carriers in the core is
important.
Aqueous Solubilization
Because the resulting QDs are coated with alkyl chains
that render solubility only in nonpolar organic solvents,
phase transfer is an essential and nontrivial step for the
QDs to be useful as biological reporters. Alternatively,
QD syntheses have been performed directly in aqueous
solution generating QDs ready to use in biological envi-
ronments,22 but these protocols rarely achieve the level of
monodispersity, crystallinity, and fluorescent efficiency as
the QDs produced in high-temperature coordinating sol-
vents. Two general strategies have been used to make hy-
drophobic QDs soluble in aqueous solution (Fig. 2): lig-
and exchange, and coating with an amphiphilic polymer.
For ligand exchange, a suspension of TOPO-coated QDs
are mixed with a solution containing an excess of a heter-
obifunctional ligand, which has one functional group that
binds to the QD surface, and another functional group that
is hydrophilic. Thereby, hydrophobic TOPO ligands are
displaced from the QD through mass action, as the new
bifunctional ligand adsorbs to render water solubility. Us-
ing this method, (CdSe)ZnS QDs have been coated with
mercaptoacetic acid and (3-mercaptopropyl) trimethoxysi-
lane, both of which contain basic thiol groups to bind to
the QD surface atoms, yielding QDs displaying carboxylic
acids or silane monomers, respectively.8,9 These methods
generate QDs that are useful for biological assays, but
ligand exchange is commonly associated with decreased
fluorescent efficiency and a propensity to aggregate and
precipitate in biological buffers. More recently it has been
shown that these problems can be alleviated by retaining
the native TOPO molecules on the surface, and covering
the hydrophobic QDs with amphiphilic polymers.16,21,61
These methods yield QDs that can be dispersed in aqueous
solution and remain stable for long periods of time due to
a protective hydrophobic bilayer encapsulating each QD
through hydrophobic interactions. No matter what method
is used to suspend the QDs in aqueous buffers, they should
be purified from residual ligands and excess amphiphiles
before use in biological assays, using ultracentrifugation,
dialysis, or filtration. Also, when choosing a water solu-
bilization method, it should be noted that many biological
and physical properties of the QDs may be affected by
the surface coating, and the overall physical dimensions
of the QDs are dependent on the coating thickness (Fig.
2). Typically, the QDs are much larger when coated with
amphiphiles, compared with those with a monolayer of
ligand.
Bioconjugation
Most water solubilization methods result in QDs covered
with carboxylic acid groups, and the QDs are negatively
charged in neutral or basic buffers. Many schemes used to
prepare QD bioconjugates rely on covalent bond forma-
tion between carboxylic acids and biomolecules (Fig. 3).9
Since the QD surface has a net negative charge, positively
charged molecules can also be used for electrostatic bind-
ing, a technique that has been used to coat QDs with cationic
avidin proteins and recombinant maltose-binding proteins
fused with positively charged peptides.24,40 Alternatively,
biomolecules containing basic functional groups, such as
amines or thiols, may interact directly with the QD sur-
face as ligands.20 If biomolecules do not innately contain
groups for direct QD binding, they may be modified to
add this functionality. For example, nucleic acids and pep-
tides have been modified to add thiol groups for binding
to QDs.1,41 Surface modification has also become modu-
lar through high-affinity streptavidinbiotin binding. QD
streptavidin conjugates are convenient for indirect bind-
ing to a broad range of biotinylated biomolecules.61 If the
QDs are intended as nonfunctional probes, then they can be
coated with inert hydrophilic polymers, such as polyethy-
lene glycol (PEG), which act to reduce nonspecific adsorp-
tion and to increase colloidal stability.7 Biocompatible QDs
are now commercially available, conjugated to a variety of
functional biological molecules, like streptavidin, biotin, or
monoclonal antibodies.
APPLICATIONS IN CELLULAR AND
MOLECULAR IMAGING
Much of our understanding of human physiology and
pathology has resulted from a strict reductionist approach
to biology. Purified biological molecules are used to predict
their behaviors in complex living cells, which are then stud-
 Engineering Luminescent Quantum Dots for In Vivo Molecular and Cellular Imaging 7

FIGURE 2. Molecular engineering to modify the physical dimensions and chemical characteristics of QDs, using common surface
coatings. The first column shows QDs with a core diameter of roughly 5.5 nm, coated with different organic compounds, drawn
to scale to demonstrate the relative sizes of the QDs and their surface coatings. The second column shows a magnified view
of each surface to clarify the molecular interactions. The third column shows dynamic light scattering data from (CdSe)ZnS QDs
coated with each of the corresponding ligands, to show their effective hydrodynamic diameters. (A) QDs coated with TOPO ligands,
which bind to the QD surface via phosphine oxide functional groups.42 These hydrophobic QDs were dissolved in chloroform. (B)
TOPO-coated QDs were ligand exchanged with mercaptoacetic acid, which binds to the QDs via thiol functional groups, causing
carboxylic acid groups to extend outward from the surface.9 These anionic QDs were dissolved in phosphate buffered saline
(PBS). (C) TOPO-coated QDs were covered with an amphiphilic anionic polymer, 40% octylamine-modified poly(acrylic acid), which
wrapped around the hydrophobic surface, intercalating with the TOPO molecules, stabilizing the surfaces through hydrophobic
interactions.61 The resulting anionic QDs were then dispersed in PBS. (D) TOPO-coated QDs were encapsulated in amphiphilic
phospholipids, a mixture of 40% 1,2-dipalmitoyl-sn-diglycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)-2000], and
60% 1,2-dipalmitoyl-glycero-3-phosphocholine.16 This QD coating was stabilized by similar effects as the polymer-coated QDs, but
after dissolution in PBS, steric stabilization and hydrophilicity of the resulting colloids was provided by long polyethylene glycol
(PEG) arms that extended outward from the QD surface. Data provided by A.M. Smith, M.N. Rhyner, and S. Nie. AU: arbitrary units.

ied in isolation to deduce their physiological roles in living
multicellular organisms. This research paradigm has been
successful, but it is also limited because purified molecules
may function differently in complex cells. For this reason, it
is desirable to study molecules of interest within their native
cellular locations, which normally requires the extraction
of cells from organismal tissues, growth of these cells in
culture, followed by sacrifice of the cells to test cellular ex-
tracts in standard and sometimes quantitative assays (e.g.,
using gel electrophoresis or immunoassays) to determine
8 SMITH et al.

FIGURE 3. Schematic illustration showing the most common methods to conjugate QDs to biological molecules such as proteins,
peptides, nucleic acids, or small organic molecules. The QD used in this illustration is rendered water-soluble using mercaptoacetic
acid, which is drawn disproportionately large to clarify its role in conjugation. Similar coupling schemes have been used with other
surface modifications as shown in Fig. 2. One exception is that direct coupling to the QD surface is much more difficult if the
coating layer thickness is large, such as for QDs coated with silica or an amphiphilic polymer. Also electrostatic attraction will only
occur if the QD has a net charge. See text for descriptions of these strategies.

the presence and/or abundance of biomolecules. This ap-
proach results in the loss of time-dependent data and spatial
information, that is, how the analyte under study changes
over time and where it exists within the heterogeneous
structures of cells and tissues. Performing assays directly on
dead cells and tissues, like immunocytochemical staining,
has been one approach to retain spatial information, but this
technique still lacks the important time element. For this
reason, it has been a long-standing goal of the biological
sciences to study living cells and tissues in real time. As
spatial information is readily viewed using an imaging tech-
nique, various types of microscopies have been developed
to image living cells with subcellular resolution, such as
electron microscopy, fluorescence microscopy, optical ab-
sorption microscopy, and magnetic resonance imaging. Of
these, the best combination of simplicity, resolution, sensi-
tivity, and robustness has been fluorescence microscopy.
Using this technique, the signal intensity from fluores-
cent molecules is used to determine the locations of fluo-
rophores and their concentrations. Typically, biomolecules
are given a fluorescent tag (an organic dye or a fluorescent
protein), which allows detection of the biomolecule using
a fluorescence microscope, or with a fluorescence imag-
ing system for whole organisms. However, signal intensity
from these tags quickly diminishes due to photobleaching,
greatly reducing the signal-to-noise ratio over time, which is
why photostable quantum dots could significantly improve
real-time fluorescence detection in living cells and organ-
isms. Also QDs can be tuned to emit in the near-infrared
spectrum (>650 nm), which could allow unmatched
sensitivity and depth penetration through biological
tissue.
 Engineering Luminescent Quantum Dots for In Vivo Molecular and Cellular Imaging 9

Living Cells blinking), demonstrating a remarkable level of sensitiv-

ity of protein detection on living cells. Protein diffusion
In 1998, two research groups simultaneously reported
was fluorescently monitored on the cellular membrane, and
that QDs could be used as fluorescent contrast agents
could even be visualized after the proteins were internal-
for labeling either living or fixed cells in culture.8,9 Fol-
ized. Dahan et al. similarly reported that glycine receptors
lowing these reports, interest in using QDs as reporters
on the membranes of living neurons could be stained with
in immunocytochemical and immunohistochemical assays
QDs coupled to an antibody fragment specific for the re-
quickly grew, due to the realization that QDs were su-
ceptor.13 Single receptors were identified, and tracking of
perior to conventional fluorescent dyes in photostability,
single QDs on the membranes allowed the calculation of
brightness, and spectral linewidth.56,61 Now it has been
protein diffusion constants. These conjugates showed supe-
established that QDantibody conjugates can be used to
rior photostability, lateral resolution, and sensitivity relative
label membrane proteins, cytosolic biomolecules, and nu-
to organic dyes.
clear antigens in fixed cells and tissues (Fig. 4).8,56,61
Detecting cellular membrane molecules on living cells
The use of QDs as live cell imaging agents has also
is very useful for many studies, but the ability to study
progressed, albeit at a slower pace due to the inherently
the complex internal machinery of living cells has been
greater difficulty of maintaining cellular viability during
a long sought-after goal. The primary method to study
staining.
proteins in their native environment has been through the
The main obstacle to QD imaging in living cells has been
expression of the protein of interest fused to a fluores-
the diffusion barrier of the cellular membrane. This is not
cent protein tag. Using this technique, the transgenic pro-
a problem if the antigen of interest resides on the plasma
tein is localized in its innate cellular location, and it may
membrane, as demonstrated by two reports in 2002. Lidke
be monitored through fluorescence microscopy. However,
et al. stained erbB/HER membrane receptors on cultured
photobleaching is commonly a problem, as is the inherent
human cancer cells using red-light emitting (CdSe)ZnS
difficulty in constructing the fusion protein. An alterna-
QDs coupled to epidermal growth factor, a small protein
tive approach is to couple fluorescent dyes to biomolecules
with a specific affinity for this receptor.36 Receptor-bound
of interest in vitro, and then introduce these conjugates
QDs were identified at the single-molecule level (single
into living cells. However, these probes must transverse
QDs may be distinguished from aggregates because the
the plasma membrane before finding their targets, and
fluorescent intensity from discrete dots is intermittent, or
also suffer from photobleaching. Using QDs in the place

FIGURE 4. Immunocytochemical stain of F-actin in fixed fibroblast cells using green quantum dots. Cells were first incubated with a
phalloidinbiotin conjugate, containing the small molecule, phalloidin, that has a specific affinity for F-actin. The biotin associated
with the F-actin was then detected by using QDstreptavidin conjugates emitting green fluorescence. This image was obtained
using an epifluorescence microscope and a 100 oil-immersion objective.
10
of dyes or proteins could alleviate photobleaching con-
cerns, and could also allow enhanced detection sensitivity
and spatial resolution. But a major problem still remains,
which is the delivery of QD probes across the plasma
membrane.
QDs have been loaded passively into cells by exploiting
the innate capacity of most cells to uptake their extracel-
lular space through endocytosis.26,31,46 This effect may be
enhanced by coupling the QDs to membrane receptors, in
part due to the higher local concentration of QDs in the
vicinity of the cellular membrane.15,31,36 However, these
methods invariably lead to sequestration of aggregated QDs
in vesicles, which are no longer free to find their intended in-
tracellular targets. Yet this is a simple way to label cells with
QDs, if specific intracellular localization is not necessary,
or if the intent is to target the endocytotic vesicles. Other
methods for internalization of QD-bioconjugates in living
cells were developed originally for the delivery of drugs
and biomolecules, and fall under two broad categories:
chemical and mechanical delivery. Chemically-mediated
delivery enhances plasma membrane translocation by us-
ing cationic lipids,14,59 membrane-permeating proteins,9 or
cationic peptides14,21,34,39 all of which have been shown
to increase cellular uptake of many different molecules
and particles (Fig. 5). Although the delivery mechanisms
are not well understood, these techniques seem to cause
QD aggregation inside of vesicles, similar to delivery by
endocytosis. Mechanical methods include microinjection
of QDs into individual cells and electroporation of cells
in the presence of QDs. Microinjection is the only tech-
nique that has been reported to deliver QDs homogeneously
into the cytoplasms of cells;14,16 however, this method is
of low statistical value, as careful manipulation of single
cells prevents the use of large sample sizes. Electropora-
tion makes use of the increased permeability of cellular
membranes under pulsed electric fields to deliver QDs,
but this method was reported to result in aggregation of
QDs in the cytoplasm.14 It is clear that if this field is to
progress, a greater understanding of this technical barrier
must be combined with new techniques to overcome this
problem.
In 2004, Derfus et al. demonstrated that QDs conjugated
to organelle-targeting peptides could specifically stain
either cellular mitochondria or nuclei, following mi-
croinjection into fibroblast cytoplasms.14 Similarly, Chen
et al. targeted peptideQDs conjugates to cellular nuclei,
using electroporation to overcome the plasma membrane
barrier.10 These schemes have resulted in organelle-level
resolution of intracellular targets for living cells, yielding
fluorescent contrast of vesicles, mitochondria, and nuclei,
but resolution on the scale of intracellular molecules has not
been reached. Although QD intracellular targeting is still
elusive, live cell labeling has already generated new assays.
The innate capacity of cells to uptake QDs via endocytosis
was utilized by Parak et al. to devise a cellular motility
SMITH et al.
FIGURE 5. Cellular uptake of QDcationic peptide conjugates.
Monolayers of living human cancer cells were incubated with
red fluorescent (CdSe)ZnS QDs conjugated to TAT peptides,
which are cationic peptides derived from an HIV-associated
membrane translocation protein. This image was obtained us-
ing confocal microscopy, a 100 oil-immersion objective, with
the focal plane near the middle of the cells. The aggrega-
tion of QDs in internal cellular structures is indicative of their
presence in endocytotic vesicles.
assay.46 The phagokinetic tracks of human breast cancer
cells were monitored through fluorescence microscopy,
as the cells transversed a QD-coated substrate. The path
traveled by each cell became dark, and the cells increased in
fluorescence as they took up more QDs. QDs have also been
used to study the particle size dependence of endocytosis,
using the fact that the surface parameters of nanoparticles
may be altered to generate specific nanoparticle sizes
(Fig. 2).45
Living Animals
Compared to studying living cells in culture, new
challenges arise with the increase in complexity to
a multicellular organism and with the accompanying
increase in size scale. Unlike monolayers of cultured cells
and thin tissue sections, tissue thickness becomes a major
concern because biological tissue attenuates most signals
used for imaging. Therefore, the major imaging techniques
used reliably for imaging whole animals rely on signals that
can transmit through thick tissue, or signals that transmit
differentially through specific types of tissue, using ultra-
sonic waves (ultrasound imaging), X-rays (computed X-ray
tomography), gamma rays (positron emission tomography),
or radio waves (magnetic resonance imaging). Optical
imaging, especially fluorescence imaging, has been used
in living animal models, but it is still limited by the poor
transmission of visible light through biological tissue. It has
been suggested that there is a near-infrared optical window
 Engineering Luminescent Quantum Dots for In Vivo Molecular and Cellular Imaging
in most biological tissue that is the key to deep-tissue
optical imaging.60 The rationale is that Rayleigh scattering
decreases with increasing wavelength, and that the major
chromophores in mammals, hemoglobin and water, have lo-
cal minima in absorption in this window. Few organic dyes
are available that emit brightly in this spectral region, and
they suffer from the same photobleaching problems as their
visible counterparts, although this has not prevented their
successful use as contrast agents for living organisms.19
One of the greatest advantages of QDs for imaging in
living tissue is that their emission wavelengths can be tuned
throughout the near-infrared spectrum by adjusting their
composition and size, resulting in photostable fluorophores
that are stable in biological buffers.33 Because visible QDs
are more synthetically advanced than near-infrared QDs,
most of the living animal studies implementing QDs have
used (CdSe)ZnS QDs that emit visible light. But even
visible QDs have shown great promise, due to their ability
to remain photostable and brightly emissive in living
organisms.
Quantum dots have been used to passively image the
vascular systems of various animal models. In a report by
Larson et al., intravenously injected green-light emitting
QDs remained fluorescent and detectable when they circu-
lated to capillaries in the adipose tissue and skin of a living
mouse.35 This report made use of two-photon excitation,
in which near-infrared light is used to excite visible QDs,
allowing for deeper penetration of excitation light, despite
strong absorption and scattering of the emitted visible light.
Lim et al. intravenously injected near-infrared (CdTe)CdSe
QDs to image the coronary vasculature of a rat heart.37 Both
of these reports utilized the native process of circulation to
deliver fluorescent probes from a single location (the tail
vein) throughout the entire vasculature. The circulatory sys-
tem is a useful means for delivering exogenous molecules,
but the delivered substances have a limited lifetime of cir-
culation due to uptake by vascularized organs. QDs, like
other nanoparticles, are taken up nonspecifically by the
reticuloendothelial system (RES), comprising phagocytotic
cells such as macrophages, mainly located in the spleen,
liver, and lymph nodes. Therefore, over time, intravenously
administered QDs will end up in these organs, unless they
have some type of targeting functionality. Also long-term
studies of the vascular system using QDs must account for
this finite lifetime in the blood stream, which could require
multiple injections of fluorescent probes. It was demon-
strated by Ballou et al. that the lifetime of QDs in the
bloodstream of mice is significantly increased if the QDs
are coated with PEG polymer chains,7 an effect that has also
been documented for other types of nanoparticles and small
molecules, in part due to decreased nonspecific adsorption
of the nanoparticles, and decreased antigenicity.54 In this
report, RES uptake of PEG-coated red-fluorescent QDs af-
ter intravenous injection in living mice was monitored in
real time using full-body imaging. Kim et al. demonstrated
11
the lymphatic system is another circulatory system that may
be used to deliver QDs to specific tissues.33 Near-infrared
QDs were intradermally injected in mice and pigs. The
QDs were engulfed by dendritic cells, which migrated to
sentinel lymph nodes, generating fluorescent contrast of
these nodes.
Cells can also be loaded with QDs in vitro, and then ad-
ministered to an organism. The fluorescent tag can be used
to identify the original cells and their progeny within the
organism. This was first demonstrated on a small organism
scale by Dahan et al. in 2001, by microinjecting QDs into
the cytoplasms of single frog embryos.16 As the embryos
grew, the cells divided, and each cell that descended from
the original labeled cell retained a portion of the fluores-
cent cytoplasm, which could be fluorescently imaged in real
time under continuous illumination. In reports by Hoshino
et al.30 and Voura et al.59 cells loaded with QDs were in-
jected intravenously into mice, and their distributions in
the animals were later determined through tissue dissec-
tion, followed by fluorescence imaging. Also Gao et al.
loaded human cancer cells with QDs, and injected these
cells subcutaneously in an immune-compromised mouse.21
The cancer cells divided to form a solid tumor, which
could be visualized fluorescently through the skin of the
mouse. With recent reports that cells may be labeled with
QDs at a high degree of specificity,34,39 it is foreseeable
that multiple types of cells may be monitored in living
organisms, and also identified using their distinct optical
codes.
The previous reports have demonstrated the capability
of using QDs in living animals, but have only demonstrated
image contrast on the level of tissues and organs. Molec-
ular imaging is the generation of image contrast due to
the molecular differences in cells, tissues, and organs. This
requires a probe that contains a molecular targeting func-
tionality to generate contrast only in locations specified
by the targeted probe. In the past, mice have been geneti-
cally altered to express fluorescent proteins in specific cell
types, and targeted fluorophores have been administrated
to animals to detect spatial differences in protein or gene
expression. Using a similar technique, Akerman et al. used
QDpeptide conjugates to target specific vascular tissues.1
Microscopic fluorescence imaging of tissue sections from
the mice demonstrated that the QDs were taken up by their
target tissues, and nonspecific RES uptake could be min-
imized via attachment of PEG to the QD surface. Whole-
animal imaging of molecular detection was realized by Gao
et al. in 2004 using red fluorescent QDs conjugated to anti-
bodies specific for antigens on a human cancer induced in a
mouse.21 Tail-vein injection of targeted QD probes resulted
in specific uptake of the probes in the cancerous tissue,
visualized using whole-body fluorescence imaging of liv-
ing mice, generating fluorescent image contrast due to the
molecular difference between normal tissue and cancerous
tissue.
12 SMITH et al.
Biocompatibility and Detection Specificity
Fluorescent detection of the events in living cells and
animals would not be very valuable if reporter molecules
significantly perturb the native biological processes. For
this reason, one of the major questions surrounding the
use of QDs in living systems is whether they are pertur-
bative or cytotoxic. The most common QD composition,
(CdSe)ZnS, contains a core of cadmium and selenium,
both of which are known to be toxic. However, almost
all studies so far have reported negligible cytotoxicity of
QDs.10,15,16,21,30,31,35,39,46,51,59 A ZnS shell, combined with
a robust organic coating, can protect the core well enough
to prevent oxidation and release of elemental cadmium
and selenium, within the time course of an assay.15 In-
deed, the only reports mentioning moderate cytotoxicity
in living cells utilized QDs coated with bifunctional cap-
ping ligands,29,30,55 which are known to be colloidally un-
stable due to constant desorption/adsorption of ligands.2
This equilibrium of surface protective molecules may have
caused exposure of the underlying semiconductor material
and QD degradation, leading to the release of cadmium and
selenium. Further research on QD surface coatings and on
protection of the underlying semiconductor crystal will be
important to ensure that QDs can be used as viable in vivo
probes.
Besides cytotoxicity, it is also important that QDs do
not interfere with other cellular or physiological processes,
either through nonspecific binding, by preventing spe-
cific binding through steric hindrance, or by changing
biomolecule functionality or diffusivity. Nonspecific bind-
ing is a problem that can be addressed through surface
coatings, as hydrophilic polymers such as PEG are known
to reduce nonspecific interactions with particles. The fact
that QDs are on the same size scale as large proteins could
alter the ability of coupled biomolecules to bind to their
targets, either by steric hindrance or by altering biomolecu-
lar functionality. The large size of QDs would also decrease
the rate of diffusion of attached biomolecules. If these prob-
lems are significant, smaller sized QD probes will need to
be designed, or small molecule organic dyes may be used
instead of QDs. It is possible that the use of QDs as targeted
imaging agents will become a nanoengineering problem, in
which the surface properties to the QD particles can be en-
gineered for biological utility, as well as increased clinical
relevance.
Detection specificity depends on the specific and non-
specific interactions of QD probes with target molecules
and the surrounding environment. Specific biomolecular
interactions are determined by the conjugated ligands such
as antibodies, peptides, nucleic acids, or small molecules.
By maximizing biomolecular binding affinity and minimiz-
ing cross reactivity, the QD probe can have highly specific
interactions. However, because QDs are nanoparticle col-
loids, they have a higher degree of nonspecific interactions
than small molecules, which makes it important to minimize
binding to background substrates. Nonspecific binding can
often be minimized using hydrophilic polymers with a high
entropy of solubilization. Polymers like PEG have been at-
tached to the surface of QD probes for reducing nonspecific
binding.
FUTURE DIRECTIONS
Quantum dots have been received as technological
marvels with characteristics that could greatly improve bio-
logical detection. However, extensive engineering and fun-
damental studies are still needed. First, QD synthesis and
post-synthesis nanoengineering must be perfected so that
reproducible, robust, and highly specific biological probes
can be generated for light emission in the visible spec-
trum, and also in the biologically important near-infrared
spectrum. Second, their use in already existing assays us-
ing conventional fluorophores should be explored to deter-
mine if they can supplant dyes and fluorescent proteins, and
whether they are potentially better labels, especially with
respect to their inherently large sizes. Third, their spectral
characteristics such as narrow emission bands, photostabil-
ity, and multicolor emission, should be tested in biological
environments to determine if these unique characteristics
are truly useful. Once these tasks have been performed, QDs
may be regarded as powerful tools for biological detection
and imaging.
There are also new biological applications of QDs that
are promising. For example, QDs are inherently strong con-
trast agents for electron microscopy due to their electron-
dense nature, relative to biological organic molecules and
water.44 Thereby, QDs are not only useful contrast agents
for macroscale imaging of entire organisms, and micron-
scale optical resolution using fluorescence microscopy, but
they can even be used to image at atomic-scale resolution
using electron microscopy. Also QDs have recently been
proposed as energy donors for photodynamic therapy, a pro-
cess in which a photosensitizer within a cell absorbs light
and transfers this absorbed energy to generate a reactive,
cytotoxic oxygen species that is able to kill cancer cells.6
This concept, combined with the potential innate cytotoxic-
ity of QDs, could result in dual-modality QD conjugates that
are both targeted imaging contrast agents and therapeutic
agents. Furthermore, simple QDs may be combined with
other nanomaterials to assemble small but complex multi-
functional machines, much in the same way that biological
proteins and other macromolecules are assembled in cells
to create multifunctional structures such as the ribosome.
These devices could be used as multimodal imaging agents,
for instance, for simultaneous image contrast for magnetic
resonance imaging and optical imaging, or even for combin-
ing specific therapeutic agents with optical imaging agents.
 Engineering Luminescent Quantum Dots for In Vivo Molecular and Cellular Imaging
ACKNOWLEDGMENTS
This work was supported by grants from the National
Institutes of Health (P20 GM072069, R01 CA108468, and
R01 GM058173), the U.S. Department of Energy Genomes
to Life Program, and the Georgia Cancer Coalition (GCC).
One of the authors (A.M.S.) acknowledges the Whitaker
Foundation for generous fellowship support.
REFERENCES
1
Akerman, M. E., W. C. W. Chan, P. Laakkonen, S. N. Bhatia, and
E. Ruoslahti. Nanocrystal targeting in vivo. Proc. Natl. Acad.
Sci. U.S.A. 99:1261712621, 2002.
2
Aldana, J., Y. A. Wang, and X. G. Peng. Photochemical insta-
bility of CdSe nanocrystals coated by hydrophilic thiols. J. Am.
Chem. Soc. 123:88448850, 2001.
3
Alivisatos, A. P. Perspectives on the physical chemistry of
semiconductor nanocrystals. J. Phys. Chem. 100:1322613239,
1996.
4
Alivisatos, A. P. Semiconductor clusters, nanocrystals, and
quantum dots. Science 271:933937, 1996.
5
Bailey, R. E., and S. M. Nie. Alloyed semiconductor quantum
dots: Tuning the optical properties without changing the particle
size. J. Am. Chem. Soc. 125:71007106, 2003.
6
Bakalova, R., H. Ohba, Z. Zhelev, M. Ishikawa, and Y. Baba.
Quantum dots as photosensitizers? Nat. Biotechnol. 22:1360
1361, 2004.
7
Ballou, B., B. C. Lagerholm, L. A. Ernst, M. P. Bruchez, and
A. S. Waggoner. Noninvasive imaging of quantum dots in mice.
Bioconjug. Chem. 15:7986, 2004.
8
Bruchez, M., M. Moronne, P. Gin, S. Weiss, and A. P. Alivisatos.
Semiconductor nanocrystals as fluorescent biological labels.
Science 281:20132016, 1998.
9
Chan, W. C. W., and S. M. Nie. Quantum dot bioconjugates for
ultrasensitive nonisotopic detection. Science 281:20162018,
1998.
10
Chen, F., and D. Gerion. Fluorescent CdSe/ZnS nanocrystal
peptide conjugates for long-term, nontoxic imaging and nuclear
targeting in living cells. Nano Lett. 4:18271832, 2004.
11
Crouch, D., S. Norager, P. OBrien, J. H. Park, and N. Pickett.
New synthetic routes for quantum dots. Philos. Trans. R. Soc.
Lond., Ser. A. 361:297310, 2003.
12
Dabbousi, B. O., J. RodriguezViejo, F. V. Mikulec, J. R. Heine,
H. Mattoussi, R. Ober, K. F. Jensen, and M. G. Bawendi.
(CdSe)ZnS core-shell quantum dots: Synthesis and character-
ization of a size series of highly luminescent nanocrystallites. J.
Phys. Chem. B 101:94639475, 1997.
13
Dahan, M., S. Levi, C. Luccardini, P. Rostaing, B. Riveau, and
A. Triller. Diffusion dynamics of glycine receptors revealed by
single-quantum dot tracking. Science 302:442445, 2003.
14
Derfus, A. M., W. C. W. Chan, and S. N. Bhatia. Intracellular
delivery of quantum dots for live cell labeling and organelle
tracking. Adv. Mater. 16:961966, 2004.
15
Derfus, A. M., W. C. W. Chan, and S. N. Bhatia. Probing the
cytotoxicity of semiconductor quantum dots. Nano Lett. 4:11
18, 2004.
16
Dubertret, B., P. Skourides, D. J. Norris, V. Noireaux, A. H.
Brivanlou, and A. Libchaber. In vivo imaging of quantum dots
encapsulated in phospholipid micelles. Science 298:17591762,
2002.
17
Efros, A. L., and A. L. Efros. Interband absorption of light in a
semiconductor sphere. Sov. Phys. Semicond. 16:772775, 1982.
13
18
Ekimov, A. I., and A. A. Onushchenko. Quantum size effect in
the optical-spectra of semiconductor micro-crystals. Sov. Phys.
Semicond. 16:775778, 1982.
19
Frangioni, J. V. In vivo near-infrared fluorescence imaging. Curr.
Opin. Chem. Biol. 7:626634, 2003.
20
Gao, X. H., W. C. W. Chan, and S. M. Nie. Quantum-dot
nanocrystals for ultrasensitive biological labeling and multicolor
optical encoding. J. Biomed. Opt. 7:532537, 2002.
21
Gao, X. H., Y. Y. Cui, R. M. Levenson, L. W. K. Chung, and S. M.
Nie. In vivo cancer targeting and imaging with semiconductor
quantum dots. Nat. Biotechnol. 22:969976, 2004.
22
Gaponik, N., D. V. Talapin, A. L. Rogach, K. Hoppe,
E. V. Shevchenko, A. Kornowski, A. Eychmuller, and H.
Weller. Thiol-capping of CdTe nanocrystals: An alternative to
organometallic synthetic routes. J. Phys. Chem. B 106:7177
7185, 2002.
23
Gerion, D., F. Pinaud, S. C. Williams, W. J. Parak, D.
Zanchet, S. Weiss, and A. P. Alivisatos. Synthesis and prop-
erties of biocompatible water-soluble silica-coated CdSe/ZnS
semiconductor quantum dots. J. Phys. Chem. B 105:88618871,
2001.
24
Goldman, E. R., E. D. Balighian, H. Mattoussi, M. K. Kuno, J.
M. Mauro, P. T. Tran, and G. P. Anderson. Avidin: A natural
bridge for quantum dot-antibody conjugates. J. Am. Chem. Soc.
124:63786382, 2002.
25
Guo, W., J. J. Li, Y. A. Wang, and X. G. Peng. Luminescent
CdSe/CdS core/shell nanocrystals in dendron boxes: Superior
chemical, photochemical and thermal stability. J. Am. Chem.
Soc. 125:39013909, 2003.
26
Hanaki, K., A. Momo, T. Oku, A. Ko-
moto, S. Maenosono, Y. Yamaguchi, and K.
Yamamoto. Semiconductor quantum dot/albumin complex is a
long-life and highly photostable endosome marker. Biochem.
Biophys. Res. Commun. 302:496501, 2003.
27
Hines, M. A., and P. Guyot-Sionnest. Synthesis and characteri-
zation of strongly luminescing ZnS-capped CdSe nanocrystals.
J. Phys. Chem.. 100:468471, 1996.
28
Hines, M. A., and G. D. Scholes. Colloidal PbS nanocrystals
with size-tunable near-infrared emission: Observation of post-
synthesis self-narrowing of the particle size distribution. Adv.
Mater. 15:18441849, 2003.
29
Hoshino, A., K. Fujioka, T. Oku, M. Suga, Y. Sasaki, T. Ohta, M.
Yasuhara, K. Suzuki, and K. Yamamoto. Physicochemical prop-
erties and cellular toxicity of nanocrystal quantum dots depend
on their surface modification. Nano Lett. 4:21632169, 2004.
30
Hoshino, A., K. Hanaki, K. Suzuki, and K. Yamamoto. Appli-
cations of T-lymphoma labeled with fluorescent quantum dots
to cell tracing markers in mouse body. Biochem. Biophys. Res.
Commun. 314:4653, 2004.
31
Jaiswal, J. K., H. Mattoussi, J. M. Mauro, and S. M. Simon.
Long-term multiple color imaging of live cells using quantum
dot bioconjugates. Nat. Biotechnol. 21:4751, 2003.
32
Kim, S., B. Fisher, H. J. Eisler, and M. Bawendi. Type-II quan-
tum dots: CdTe/CdSe(core/shell) and CdSe/ZnTe(core/shell)
heterostructures. J. Am. Chem. Soc. 125:1146611467,
2003.
33
Kim, S., Y. T. Lim, E. G. Soltesz, A. M. De Grand, J. Lee, A.
Nakayama, J. A. Parker, T. Mihaljevic, R. G. Laurence, D. M.
Dor, L. H. Cohn, M. G. Bawendi, and J. V. Frangioni. Near-
infrared fluorescent type II quantum dots for sentinel lymph
node mapping. Nat. Biotechnol. 22:9397, 2004.
34
Lagerholm, B., M. Wang, L. Ernst, D. Ly, H. Liu, M. Bruchez,
and A. Waggoner. Multicolor coding of cells with cationic pep-
tide coated quantum dots. Nano Lett. 4:20192022, 2004.
35
Larson, D. R., W. R. Zipfel, R. M. Williams, S. W. Clark, M.
P. Bruchez, F. W. Wise, and W. W. Webb. Water-soluble quan-
14 SMITH et al.
tum dots for multiphoton fluorescence imaging in vivo. Science
300:14341436, 2003.
36
Lidke, D. S., P. Nagy, R. Heintzmann, D. J. Arndt-Jovin, J.
N. Post, H. E. Grecco, E. A. Jares-Erijman, and T. M. Jovin.
Quantum dot ligands provide new insights into erbB/HER
receptor-mediated signal transduction. Nat. Biotechnol. 22:198
203, 2004.
37
Lim, Y. T., S. Kim, A. Nakayama, N. E. Stott, M. G. Bawendi,
and J. V. Frangioni. Selection of quantum dot wavelengths for
biomedical assays and imaging. Mol. Imag. 2:5064, 2003.
38
Mahtab, R., J. P. Rogers, C. P. Singleton, and C. J. Mur-
phy. Preferential adsorption of a kinked DNA to a neutral
curved surface: Comparisons to and implications for nonspecific
DNA-protein interactions. J. Am. Chem. Soc. 118:70287032,
1996.
39
Mattheakis, L., J. Dias, Y. Choi, J. Gong, M. Bruchez, J. Liu,
and E. Wang. Optical coding of mammalian cells using semi-
conductor quantum dots. Anal. Biochem. 327:200208, 2004.
40
Mattoussi, H., J. M. Mauro, E. R. Goldman, G. P. Anderson, V.
C. Sundar, F. V. Mikulec, and M. G. Bawendi. Self-assembly of
CdSeZnS quantum dot bioconjugates using an engineered re-
combinant protein. J. Am. Chem. Soc. 122:1214212150, 2000.
41
Mitchell, G. P., C. A. Mirkin, and R. L. Letsinger. Programmed
assembly of DNA functionalized quantum dots. J. Am. Chem.
Soc. 121:81228123, 1999.
42
Murray, C. B., D. J. Norris, and M. G. Bawendi. Synthesis and
characterization of nearly monodisperse CdE (E = S, Se, Te)
semiconductor nanocrystallites. J. Am. Chem. Soc. 115:8706
8715, 1993.
43
Nirmal, M., and L. Brus. Luminescence photophysics in semi-
conductor nanocrystals. Acc. Chem. Res. 32:407414, 1999.
44
Nisman, R., G. Dellaire, Y. Ren, R. Li, and D. P. Bazett-Jones.
Application of quantum dots as probes for correlative fluores-
cence, conventional, and energy-filtered transmission electron
microscopy. J. Histochem. Cytochem. 52:1318, 2004.
45
Osaki, F., T. Kanamori, S. Sando, T. Sera, and Y. Aoyama. A
quantum dot conjugated sugar ball and its cellular uptake on the
size effects of endocytosis in the subviral region. J. Am. Chem.
Soc. 126:65206521, 2004.
46
Parak, W. J., R. Boudreau, M. Le Gros, D. Gerion, D. Zanchet,
C. M. Micheel, S. C. Williams, A. P. Alivisatos, and C. Larabell.
Cell motility and metastatic potential studies based on quantum
dot imaging of phagokinetic tracks. Adv. Mater. 14:882885,
2002.
47
Peng, X. G. Mechanisms for the shape-control and shape-
evolution of colloidal semiconductor nanocrystals. Adv. Mater.
15:459463, 2003.
48
Peng, Z. A., and X. G. Peng. Formation of high-quality CdTe,
CdSe, and CdS nanocrystals using CdO as precursor. J. Am.
Chem. Soc. 123:183184, 2001.
49
Peng, X. G., J. Wickham, and A. P. Alivisatos. Kinetics of IIVI
and IIIV colloidal semiconductor nanocrystal growth: "Focus-
ing" of size distributions. J. Am. Chem. Soc. 120:53435344,
1998.
50
Pietryga, J., R. Schaller, D. Werder, M. Stewart, V. Klimov,
and J. Hollingsworth. Pushing the band gap envelope: Mid-
infrared emitting colloidal PbSe quantum dots. J. Am. Chem.
Soc. 126:1175211753, 2004.
51
Pinaud, F., D. King, H. P. Moore, and S. Weiss. Bioactivation
and cell targeting of semiconductor CdSe/ZnS nanocrystals with
phytochelatin-related peptides. J. Am. Chem. Soc. 126:6115
6123, 2004.
52
Qu, L. H., and X. G. Peng. Control of photoluminescence
properties of CdSe nanocrystals in growth. J. Am. Chem. Soc.
124:20492055, 2002.
53
Qu, L. H., Z. A. Peng, and X. G. Peng. Alternative routes toward
high quality CdSe nanocrystals. Nano Lett. 1:333337, 2001.
54
Roberts, M., M. Bentley, and J. Harris. Chemistry for peptide
and protein PEGylation. Adv. Drug Delivery. Rev. 54:459476,
2002.
55
Shiohara, A., A. Hoshino, K. Hanaki, K. Suzuki, and K. Ya-
mamoto. On the cyto-toxicity caused by quantum dots. Micro-
biol. Immunol. 48:669675, 2004.
56
Sukhanova, A., M. Devy, L. Venteo, H. Kaplan, M. Artemyev,
V. Oleinikov, D. Klinov, M. Pluot, J. H. M. Cohen, and I.
Nabiev. Biocompatible fluorescent nanocrystals for immunola-
beling of membrane proteins and cells. Anal. Biochem. 324:60
67, 2004.
57
Talapin, D. V., A. L. Rogach, A. Kornowski, M. Haase,
and H. Weller. Highly luminescent monodisperse CdSe and
CdSe/ZnS nanocrystals synthesized in a hexadecylamine-
trioctylphosphine oxide-trioctylphospine mixture. Nano Lett.
1:207211, 2001.
58
Talapin, D. V., A. L. Rogach, I. Mekis, S. Haubold, A.
Kornowski, M. Haase, and H. Weller. Synthesis and surface
modification of amino-stabilized CdSe, CdTe and InP nanocrys-
tals. Colloids Surf. A 202:145154, 2002.
59
Voura, E., J. Jaiswal, H. Mattoussi, and S. Simon. Tracking
metastatic tumor cell extravasation with quantum dot nanocrys-
tals and fluorescence emission-scanning microscopy. Nat. Med.
10:993998, 2004.
60
Weissleder, R. A clearer vision for in vivo imaging. Nat. Biotech-
nol. 19:316317, 2001.
61
Wu, X. Y., H. J. Liu, J. Q. Liu, K. N. Haley, J. A. Treadway, J. P.
Larson, N. F. Ge, F. Peale, and M. P. Bruchez. Immunofluores-
cent labeling of cancer marker Her2 and other cellular targets
with semiconductor quantum dots. Nat. Biotechnol. 21:4146,
2003.
62
Yu, M. W., and X. G. Peng. Formation of high-quality CdS
and other IIVI semiconductor nanocrystals in noncoordinating
solvents: Tunable reactivity of monomers. Angew. Chem. Int.
Ed. 41:23682371, 2002.
63
Yu, W. W., L. H. Qu, W. H. Guo, and X. G. Peng. Experimental
determination of the extinction coefficient of CdTe, CdSe, and
CdS nanocrystals. Chem. Mater. 15:28542860, 2003.
64
Zhong, X. H., Y. Y. Feng, W. Knoll, and M. Y. Han. Alloyed
Znx Cd1x S nanocrystals with highly narrow luminescence spec-
tral width. J. Am. Chem. Soc. 125:1355913563, 2003.
65
Zhong, X. H., M. Y. Han, Z. Dong, T. J. White, and W. Knoll.
Composition-tunable Znx Cd1x Se nanocrystals with high lu-
minescence and stability. J. Am. Chem. Soc. 125:85898594,
2003.