Detection and classification of single viruses
Because of the devastating potential for rapid
infection from a small amount of biological agents, biological weapons
are likely candidates for future terrorist attacks. Through recent
events we have already experienced the magnitude of the threat associated
with biological agents. Examples are the disruption to our government
caused by Anthrax spores or the economic harm caused by the SARS
outbreak. Warfare viruses are especially dangerous because there
are no existing cures. Early detection is the only defense against
this threat, but because of the very small size of viruses it is
extremely challenging to detect them. Today, this requires complex
and expensive equipment and highly trained personal, and due to
this complexity and cost, these means cannot be made widely available.
We believe that optical sensors can be
developed to provide an accurate, simple, and affordable way of
detecting small biological agents. There are many challenges remaining
to build this sensor, but our studies show that a real-time single-virus
detector is possible.
Viruses and other biological species can be characterized
according to size, shape, and optical / spectroscopic properties.
These properties allow them to be distinguished from other biological
species and from other particulates such as dust particles.
Optical force for virus size and shape assessment
We developed a technique, where size and shape
of a nanometer-size particle (virus particles can be considered
as nanoparticles) are measured by detecting the optical gradient
force. While being carried by fluid inside a flow-cell (see figure
below), nanoparticle motion is perturbed by a strongly focused laser
beam due to the optical force. Back-scattered light is detected
by a photodiode which is integrated into a feedback loop with the
modulator, which prevents clumps of viruses or other large particles
from being trapped and thus from blocking the detector. The light
scattered in forward direction is used to track particle position
with respect to the focus.
Schematic of the optical gradient forces sensor
The particle motion perturbation is reflected
in the temporal asymmetry of the detector signal. Such signal asymmetry
yields information about the force exerted on the particles by the
laser focus. Large particles, which experience strong forces while
moving trough the laser focus, correspond to a highly asymmetric
detector signal, whereas small particles pass the laser focus almost
unperturbed thus rendering a symmetric signal.
Distribution for a sample of mixed 50nm and 100nm
radius polystyrene beads in water based on force measurement
The outlined scheme is well suited for the detection
of ellipsoidal particles. In addition to the optical force, the
laser focus exerts a torque on the oddly shaped particles. When
ellipsoidal particles move in random orientations towards the laser
focus, the focus forces the particles to align in a certain direction.
A change in orientation results in a change of the scattered field
which can be seen as the modulation of the differential signal from
the quadrant detector. The modulation frequency therefore provides
information about eccentricity of the particles.
Background-free detection of viruses
Current optical detection methods which are well
developed for single micrometer size particles, cannot be applied
to to nanoparticles due to a strong signal dependence on particle
size. Typically, such sensors consist of a light source which illuminates
a sample volume of an aerosol or a liquid flow containing the particles
of interest. An off-axis detectors measures power of scattered light.
The latter is a function of particle properties such as size, concentration,
and optical density. In the tens of nanometers size regime particles
act as dipoles, therefore the power of scattered light is proportional
to the sixth power of particles size. Lowering the detection size
limits for the existing detectors places an impossible requirement
on noise optimization. Therefore, a signal which has a weaker particle
size dependence can allow access to smaller particles.
The detection scheme is schematically shown in
the figure below. Using the electroosmotic effect, a particle solution
is transported through a microfluidic channel. A laser beam is split
by a beamsplitter into two perpendicular paths. One path serves
as a reference for later interferometric recombination and the other
path is focused with an objective lens into a single pre-selected
nanochannel. The lateral dimensions of the nanochannels are comparable
to the size of the laser focus ensuring that no more than one particle
crosses the focus at any time. The backscattered light from a particle
passing through the laser focus is collected with the same objective
and is then recombined with the reference beam and directed onto
a split photodetector. The power of the reference beam can be arbitrarily
Schematic for the background-free interferometric single particle
The intensity distribution on the detector surface
is calculated as
where subscript r denotes reference beam and s denotes
scattered field. The signal measured by the split detector corresponds
to the difference between two halves of the detector surface normalized
by the total power incident on the detector. As the result, the
signal is proportional to the interferometric term only. The interferometric
term is proportional to the scattered field amplitude, which
has a weaker particle size dependence, compared to the conventional
sensors (third power of particle size).
(a) Data acquired for a mix of 15nm and 50nm polystyrene beads in
water and (b) for a mix of 7nm and 20nm gold particles in water.
Live virus detection
The above principle of interferometric particle
detection can be applied to a sample containing live viruses:
Data acquired Influenza A virus and 100nm polystyrene
beads in HEPES
Raman scattering for chemically specific virus
Different to elastic light
scattering, inelastic light scattering results in an energy
change of the scattered radiation. Examples of inelastic scattering
are fluorescence and Raman scattering. In the latter, an incident
photon interacts with the molecular vibrational modes of a sample.
As a result, the spectrum of the scattered radiation contains vibrational
lines that provide a unique and characteristic fingerprint for the
chemical composition of a sample. Thus, Raman scattering can provide
information about the inner protein structure of a virus and can
be used to recognize different strains of the same virus family.
In the current project, we combine nanoparticle sizing methods with
inelastic light scattering in order to add chemical specificity
to the detection process. Such sensor would have the property of
being able to sense smallest amounts of dangerous viruses, for example
near buildings, luggage, or waste water. A single virus will be
classified according to its size, shape, and its optical and spectroscopic