WO2009150526A2

WO2009150526A2 - Photothermal detection system

Info

Publication number: WO2009150526A2
Application number: PCT/IB2009/005935
Authority: WO
Inventors: Herman Pieter Spaink; Thomas Schmidt; Michel Alain Gaston Julien Orrit; Florian Markus Kulzer
Original assignee: Universiteit Leiden; Stichting Voor De Technische Wetenschappen
Priority date: 2008-06-13
Filing date: 2009-06-15
Publication date: 2009-12-17
Also published as: GB0810893D0; WO2009150526A3

Abstract

A photothermal system for performing high throughput single particle detection, the system comprising: a first optical path for directing a probe laser beam (PB) and a heating laser beam (MHB) towards a target or sample area (62); and a second optical path for directing the probe laser beam from the sample to a detector for detecting phase changes (ΔΦ) induced in the probe beam (PB) by the photothermal effect due to local heating of the sample by the heating beam (MHB), wherein the first optical path includes one or more optical elements (60a, 60b) arranged to define a detection volume of more than 10 femtolitres.

Description

Photothermal Detection System

Field of the Invention

The present invention relates to a photothermal detection system and apparatus for performing high-throughput single-particle detection.

Background of the Invention

Conventionally, detection of organic molecules, biomolecules and similar species is carried out by optical microscopy, which uses fluorescent labels to identify a target. However, optical microscopy suffers from blinking and bleaching problems widely associated with the use of such fluorescent labels, and additional problems can arise from the toxicity of many fluorescent dyes. Recently, photothermal microscopy has been introduced as a new method for detecting organic materials. This technique uses a refractive index gradient induced by heating via plasmonresonant absorption of a pump beam as a means for scattering a probe beam. Particles that exhibit this type of resonance include metal particles, for example gold nanoparticles, as well as semiconductor quantum dots. Metal nanoparticles have absorption cross sectional areas that are three to four orders of magnitude greater than that of common fluorescent labels. Because of this, high strength photothermal microscopy signals can be obtained using metal labels, thereby allowing increasingly smaller particles to be detected via photothermal microscopy, which in turn allows increasingly smaller particles to be employed as labels of, for example, sensitive biological components.

In photothermal detection, a nanoparticle is irradiated with a pump laser beam with an optical frequency in the vicinity of the plasmon resonance. The photon energy absorbed is dissipated as heat to the surrounding environment. The detected signal is provided by scattering of a secondary probe beam, due to the gradient of refractive index that is generated by the local heating. Thus, the scattered field comes from the heated region around the particle and not only from the particle itself, as in other scattering techniques. The signal intensity is proportional to the particle absorption cross-section area and, since this quantity scales with the particle volume, the signal decrease with the particle size is not too drastic, which allows for observation of smaller objects. For instance, using photothermal heterodyne imaging with high numerical aperture optics, it is possible to individually detect gold particles with diameters down to 1.4 nm, which corresponds to a metal cluster of only 67 atoms, see Berciaud, S.; Cognet, L.; Blab, G. A.; Lounis, B.

Photothermal Heterodyne Imaging of Individual Nonfluorescent Nanoclusters and

Nanocrystals Phys. Rev. Lett. 2004, 93, 257402. WO2004/025278 describes a high-resolution system and method for photothermal imaging of individual metal particles. In this, a reference laser beam and a probe laser beam including a heating laser beam are passed through a medium, in which tiny metal particles are immersed. The reference laser beam and the probe laser beam are reflected after passing through the medium to return along their original optical paths for detection. The probe laser beam undergoes phase changes when passing through the medium induced by a photothermal effect due to local heating of the medium via the heating beam. In contrast, the reference laser beam does not undergo phase change. The phase differences between the probe and reference lasers can be detected through differential photothermal interference to allow the tiny metal particles to be imaged as optical labels. To achieve a high resolution, a microscope-type focusing unit having high numerical aperture (NA) value of 1.4 is used for focusing the reference laser beam and the probe laser beam to very small focal volume.

WO 2006/013272 discloses another system and method for photothermal imaging of nano-objects in a light refracting medium. This describes the passing of a probe laser beam including a heating laser beam through a medium and combining the passed probe laser beam with a reference laser beam to image metal nanoparticles included in the medium as optical labels by performing differential photothermal interference. The device disclosed has a structure in which the probe laser beam is reflected after passing through the medium to return along its original path for detection. It also has a microscope type- focusing unit having high numerical aperture (NA) value of 1.4 for performing high- resolution detection and tracking of individual nanoparticles.

Summary of the Invention

According to an aspect of the present invention, there is provided a system for performing high throughput single particle detection, the system comprising: a first optical path for directing a probe laser beam and a heating laser beam towards a target or sample area; and a second optical path for directing the probe laser beam from the sample to a detector for detecting phase changes induced in the probe beam by photothermal effect due to local heating of the sample by the heating beam, wherein the first optical path includes one or more optical elements arranged to define a detection volume of more than 10 femtolitres. Preferably the detection volume is more than 20 femtolitres. More preferably, the detection volume is more than 30 femtolitres, ideally more than 40 femtolitres, for example 50 femtolitres. According to another aspect of the present invention, there is provided a system for performing high throughput single particle detection, the system comprising: a first optical path for directing a probe laser beam and a heating laser beam towards a target or sample area; and a second optical path for directing the probe laser beam from the sample to a detector for detecting phase changes induced in the probe beam by photothermal effect due to local heating of the sample by the heating beam, wherein the first optical path includes one or more optical elements arranged to have a numerical aperture of less than 1. Preferably, the numerical aperture is less than 0.8, for example 0.5. This allows substantially larger detection volumes than previously used in photothermal detection systems.

According to another aspect of the present invention, there is provided a system for performing high throughput single particle detection, the system comprising: a first optical path for directing a probe laser beam and a heating laser beam towards a target or sample area; and a second optical path for directing the probe laser beam from the sample to a detector for detecting phase changes induced in the probe beam by photothermal effect due to local heating of the sample by the heating beam, wherein the first and second optical paths are non-coincident.

One or more optical elements may be provided on the first optical path arranged to have a numerical aperture of less than 1. Preferably, the numerical aperture is less than 0.8, for example 0.5.

One or more optical elements may be provided on the first optical path to define a detection volume of more than 10 femtolitres. Preferably, the detection volume is more than 20 femtolitres. More preferably, the detection volume is more than 30 femtolitres, ideally more than 40 femtolitres, for example 50 femtolitres.

The detector may use a reference laser beam to detect phase changes in the probe laser beam after it has passed through the sample. The reference beam may be derived from the same source as the probe laser beam.

According to another aspect of the present invention, there is provided a screening process comprising: screening a target substance for active particles to which labels are attached, for example metal nanoparticles, using a system for performing high throughput single particle detection in accordance with the other aspects of the invention, and isolating particles when a positive detection result is obtained. When the target substance is a liquid or solution, the process may further comprise separating a sample volume of the liquid when a positive detection result is obtained, collecting together all separated sample volumes that had positive detection results, diluting the collected sample volumes and re-screening the collected sample volumes in order to enrich the solution with active particles. These steps may be repeated as required, possibly with varying conditions to increase specificity, for example by successively adding competitive binding agents in later stages of the process.

Brief Description of the Drawings

Various aspects of the invention will now be described by way of example only and with reference to the embodiments shown in the accompanying drawings in which:

Figure 1 is a schematic diagram of a forward and backward mode photothermal detection apparatus for performing high throughput single particle detection; Figure 2 is an illustration of the fields that contribute to the photothermal effect;

Figure 3 shows a comparison between the detectable volumes for low-NA photothermal and high-NA fluorescence-based detection;

Figure 4 is a series of graphs showing results of screening experiments performed using the photothermal detection apparatus of Figure 1 ; Figure 5 is a flowchart of a typical screening process that can be performed by using the apparatus of Figure 1 ;

Figure 6 is a top view of a channel layout in a microfluidic device; and

Figure 7 is a photothermal trace showing detection of a bacteriophage virus labelled with 20 nm gold particles in aqueous environment obtained using the forward mode photothermal detection apparatus of Figure 1.

Specific Description of the Drawings

In accordance with the invention, the photothermal detector uses a detection volume that is larger, by for example more than two orders of magnitude, than the volumes typically used in fluorescence-based high-sensitivity detection. Although this reduces the detection efficiency, the strength of the photothermal signal still allows for a reasonable signal-to- noise ratio without requiring long integration times. This extended detection volume allows for high-throughput screening of single particles, which has not previously been possible.

Figure 1 shows a photothermal detection apparatus 50 for performing high throughput single particle detection. This is arranged to define a detection volume of more than 10 femtolitres, and ideally more than 20 femtolitres. This apparatus can be used in two modes - forward scattering detection and reverse scattering detection. The apparatus 50 has a Helium-Neon laser 52 (wavelength=632.8 nm; spectrally well-separated from the heating beam) providing a probe beam PB. On the optical path of the probe beam PB there is a polarizing beam splitter 54 that allows light with the specific linear polarization of the probe beam PB to pass through it.

On a first optical path from the beam splitter 54, there is a quarter waveplate 56, which converts the probe beam PB to circular polarization, a first dichroic mirror 58a that is transparent to the probe beam PB, a first objective lens 60a (NA = 0.4) for focusing the probe beam onto a target sample area, a sample 62 including metal nanoparticles P arranged on a 3-axis piezo scanner 64, a second objective lens 60b (NA = 0.42) for collecting light from the sample 62, a second dichroic mirror 58b, a red pass filter 66a and a forward mode detector 68a. On another optical path from the beam splitter 54 there is a band pass filter 66b and a reverse mode detector 68b. The band pass filters 66a and 66b are identical and are selected to pass wavelengths that correspond to the Helium-Neon laser 52, while blocking the wavelength of the heating beam.

To heat the sample 62, an Argon ion laser 70 (wavelength=514.5nm) is used to generate a heating beam HB. This is time modulated with a frequency of around 1MHz by an acousto-optical modulator (AOM) 72 to provide a modulated heating beam MHB that facilitates heterodyne detection. The modulated heating beam MHB is directed onto the first dichroic mirror 58a. The first dichroic mirror 58a is selected to be reflective to the modulated heating beam MHB and the second dichroic mirror 58b is selected to be transmissive to the modulated heating beam MHB.

In forward mode, the probe beam PB passes through the polarizing beam splitter 54, the quarter waveplate 56, and the first dichroic mirror 58a where the modulated heating beam MHB is superimposed. The first dichroic mirror 58a is aligned to ensure that the probe beam and the time modulated heating beam MHB propagate in the same direction. After this, both the probe beam PB and the modulated heating beam MHB pass through the first objective lens 60a where they are focussed onto the sample 62. While passing through the sample 62, the probe beam PB undergoes slight phase changes ΔΦ induced by the photothermal effect due to local heating resulting from the modulated heating beam MHB. After passing through the sample, the probe beam has two components - a first relatively large component that is substantially unaffected by the sample and can be used as a reference (reference beam RB), and a second smaller component that is due to scattering caused by the photothermal effect stimulated by the modulated heating beam MHB.

After passing through the sample 62 some of the probe beam is reflected from the sample carrier, and the rest propagates forward to pass through to the second objective lens 60b. It is this forward propagating light that is used in the forward mode detection. At the second objective lens 60b, the forward mode beam is collected and passed to the second dichroic mirror 58b. At the second dichroic mirror 58b, the modulated heating beam MHB passes straight through whereas the two-component probe beam is reflected to pass through the red pass (or bandpass) filter 66a to the forward mode detector 68a. Thus the modulated heating beam MHB and the probe beam PB are separated from each other and the phase changed probe beam PB(ΔΦ) is incident on the forward mode detector 68a, together with the reference beam RB.

The intensity of the heating beam is time-modulated and so is the thermally induced response of the refraction index profile around the nanoparticle. To retrieve the thermally induced signal, the signal received at the forward mode detector 68a is processed with reference to the AOM 72 and using the amplifier 74. Lock-in amplification is used to retrieve the component of the signal beating at the modulation frequency. Other components that are constant or beat at high frequencies in relation to the modulation frequency are filtered out electronically. This is a process that is generally referred to as heterodyne detection.

In reverse mode, the probe beam PB passes through the polarizing beam splitter 54, the quarter waveplate 56, the first dichroic mirror 58a where the modulated heating beam MHB is superimposed. The first dichroic mirror 58a is aligned to ensure that the probe beam and the modulated heating beam MHB propagate in the same direction. After this, both the probe beam PB and the modulated heating beam MHB pass through the first objective lens 60a, and the sample 62. While passing through the sample 62, the probe beam PB undergoes slight phase changes ΔΦ and scattering, induced by the photothermal effect due to local heating resulting from the modulated heating beam MHB. At the sample, some of the light will be reflected from the sample carrier to form a reverse beam RB. It is this reflected light that is used for the reverse mode detection.

Due to the photothermal effect, the reverse beam RB has two components - a first relatively large component, reflected at an interface, that is substantially unaffected by the sample and can be used as a reference, and a second smaller component that is due to scattering caused by the photothermal effect stimulated by the modulated heating beam MHB. At the first dichroic mirror 58a only the back-scattered probe beam RB passes through. After this the reverse beam RB passes through the quarter waveplate 56, which changes the polarization from circular to a linear polarization. At this stage, the polarisation is such that the reverse beam is reflected from the beam splitter 54 to pass through the red pass filter 66b and then to the reverse mode detector 68b. The received signal is passed to the lock-in amplifier 74, where it is processed with reference to the AOM 72 to isolate the scattering component of the reverse beam RB which is generated with respect to metal nanoparticles P in the sample 62, while suppressing other contributions such as scattering from interfaces or impurities.

The detector of Figure 1 can be used in either the reverse or forward modes. Typically, the decision on which mode to use depends on the nature of the samples under investigation. In some cases, measurements could be taken in both modes to compare or validate results, or even used simultaneously.

Figure 2 shows the fields contributing to the photothermal signal. These are the scattered field by the thermally-induced refraction index profile E_{sc nW}, the scattered field by the particle itself E_sc,p and the reflected field by the bottom interface of the measurement cell E_ref. In the case of the reverse mode, the two reference fields that act as local oscillators are the scattered field by the particle itself E_SCιP, and the fraction of probe field reflected at the bottom glass interface £_rθf.

In contrast to prior art arrangements that have numerical apertures (NA) of around 1.3, the photothermal detector of Figure 1 has an effective numerical aperture of around 0.4.

This reduction in NA increases the focal volume of the probe beam in the sample and so increases the detectable sample volume to more than 10 femtolitres, thereby allowing high throughput detection of individual particles. In practice, single particle detection has been demonstrated in detection volumes of 40 femtolitres and 50 femtolitres. The fact that single particles can still be detected in a relatively large volume using photothermal detection has not been previously appreciated.

As well as using a relatively large detection volume, throughput can be further increased by using a scanning arrangement. This could involve either scanning the probe beam over the sample or moving the sample relative to the beam. Both of these techniques are well known in the art as so will not be described in detail. The dimensions of the detectable sample volume (v) using the photothermal detection apparatus 10 are proportional to 1/NA in each of the two lateral directions and to (1/NA)² along the optical axis of the PB leading to an overall enlargement of the detectable sample volume of (1/NA)⁴. That is, Vd oc (— )⁴

^K NA^}

Figure 3 shows a comparison of the detection volumes for high and low NA values in photothermal detection apparatus. The Figure shows the half-maximum isosurface for the detected signal as a solid oval; the analogous volume for high-NA microscopy is indicated by the dashed outline. As can be seen from Figure 3, the volume of samples detectable in a low NA apparatus shown by the solid oval is far greater than that of a high NA apparatus, which is indicated by a dashed line. The photothermal detection apparatus of the present invention utilizes a detection volume more than two orders of magnitude larger than that conventionally used.

Figure 4 shows time traces of photothermal signals for 80-20 nm gold particles measured using the detector of Figure 1 , which has a numerical aperture of 0.4 and so a large detection volume. Even for 20nm particles the bursts are strong enough to be clearly discriminated against the instrumental noise. This clearly demonstrates the ability of the present invention to detect individual nanoparticles while processing higher volumes of samples than conventional systems.

Because heterodyne detection is sensitive to phase, when the scattered and reference fields are in-phase the signal builds up to positive amplitudes, while when they are out-of- phase the signal drops off to negative amplitudes, going through nodes at quarter-phase shifts. The contribution of an additional in-phase local oscillator diminishes the visibility of the fringe pattern created by the interference and, therefore, the bursts have predominantly positive amplitude, but within the envelope of each burst fast fluctuations that are the signature of the interference pattern can be identified.

In the detector of Figure 1 , high throughput single particle detection can be achieved, in contrast to conventional high NA detectors, as illustrated below, where we compare high- NA fluorescence detection (first row), fluorescence detection with lower NA optics, which requires a longer detection time (second row), and the capabilities of low-NA photothermal detection (third row):

The invention further includes a detection technique to improve the throughput in screening applications using the photothermal detection apparatus of Figure 1. These are applications whose goal it is to find and separate active biochemical species (those that bind to a target) from much more numerous inactive ones (which do not bind to the same target). This is well known to be a "needle-in-a-haystack" problem. However, using the photothermal detection apparatus of the invention active particles can be detected in larger detection volumes while processing much higher total volumes of samples than using conventional techniques.

Generally speaking, the screening process proceeds as set out in Figure 5. This process can be performed using the system of Figure 1. First a large volume (V) of suspension or solution containing a few active compounds, characterized by efficient binding to the biochemical target that has been attached to the nanoparticles, and many inactive ones is prepared 100. Then each detection volume (v) in volume V is screened 102. This requires a time T=t*V/v, where t is the detection time for each detection volume v. The larger the detection volume v, the shorter the screening time T. Since the detection technique allows for a low NA optics, the detection volume can be increased by up to 1000 times compared to approaches that require a high NA, and this allows for huge gains in processing times.

There are different possible ways to screen. If all particles are immobilized on a surface, the beam may be scanned across the surface. For one application of the invention (phage display libraries), it is better to keep the phages in solution. In this case, it is easier to flow the whole solution across a small channel (e.g. in a microfluidic device) adapted to the detection volume, i.e. a small channel that has similar dimensions to the detection volume. If clogging issues arise, a broader channel may be used and the probe and heating beams may be scanned across it to probe all flowing detection volumes in the typical flow time. Alternatively, the channel may be moved relative to the beams.

Once a positive signal has been detected from the current detection volume 104, this volume is separated from the rest of the flowing solution 108. There are various microfluidic options for this, but in all cases, much larger separated volumes than the detection volume itself will be obtained. Therefore, there will be many false positives (inactive particles accidentally diverted along with the active particle). If no positive signal is detected the screening continues onto subsequent detection volumes 106 until either a positive detection signal is obtained or until all of the prepared solution has been screened 110.

The separated volumes are then diluted by a large factor 1 14, for example 1000, and the screening is repeated 116. The concentration of the inactive particle will therefore be decreased 1000 times after this second step, while all the active particles will still be there. By repeating this stage several times, the solution can be enriched in active particles up to any arbitrary extent, allowing extremely rare species with favourable binding properties to be isolated. This stage repeats until the prepared solution has been screened at a desired level of dilution 112. This process is naturally open to many variations. For example, the conditions for binding labels between these enrichment steps may also be varied.

Detection volume and detection time depend on the label in use. It is easier (and therefore faster for a given detection volume v, or doable with a larger detection volume for a given detection time t) to detect a big gold nanoparticle that has a 200 nm diameter than a small one that has a diameter of 5 nm. In the optimization of the ratio t/v, photothermal detection has two crucial advantages over fluorescence: (i) gold labels do not bleach, so the iterative process (scanning-diluting-scanning) may be applied as many times as required, without losing the targets: this is not the case for fluorescent labels or fluorescent beads, due to the inevitable occurrence of photobleaching in the early stages of the sorting process; and (ii) photothermal detection is extremely sensitive, so that the detection volume can be increased very much without losing single-particle sensitivity. This allows for high throughput. In fluorescence, to achieve single-label detection, one must have a very small confocal volume of 0.1 fl_. That leads to very long screening times for large volumes because the detection time cannot be much shorter than 1 ms. Figure 6 shows a top view of a microfludic device 80 which may be used in the process described above. As shown in Figure 6, a plurality of channels 92 is formed in the microfludic device 80. These can communicate with input/output channels 94 through which samples pass for sorting. The channels 92 have a depth of approximately 15 μm (matched to the dimension of the detection volume in that direction) and a width in the range of approximately 10-100 μm. Before the output channel, a detection region is defined where the probe beam of the photothermal detection apparatus 50 passes through a sample. Accordingly, the sample may be sorted according to the requirements of a user and output as desired. To facilitate this, at least one input channel and at least two output channels, for example 3 channels 94 are formed to allow the sample to be selectively flowed out of one of the output channels 94 in order to separate the active particles from the entire solution.

In addition to performing experiments showing forward mode detection capability of nanoparticles in water, experiments were done in which M13 bacteriophage virus was labelled with 20 nm gold particles at the coating protein P8 in aqueous environment. The results of this are shown in Figure 7 in which peaks representing detected nanoparticles can clearly be seen. The heating powers used for detection of this sample corresponded to an average temperature increase of about 3 Kelvin around the 20 nm gold particles. However, the surface temperature and the joint effect of several labels could correspond to higher temperatures (in the order of some tens of Kelvin) that could be prejudicial for biological systems. On the other hand, the residence times of the detected objects in the observation volume range from some tens to hundreds of milliseconds, which could be short enough to avoid substantial heating damage. A possible alternative to minimize heating damage is to reduce the heating power and to compensate the signal loss with an increase in the integration time at the expense of some time resolution in the correlation measurements, which could be afforded in, for example, the above case of the bacteriophage sample.

When using the photothermal detection apparatus the effect of temperature induced by the modulated heating beam MHB may have to be taken into account, especially if the sample includes biological components or other highly temperature sensitive materials. By reducing the power of the modulated heating beam MHB the temperature change ΔT can be reduced as required. However, this will reduce the signal strength and as such any reduction in temperature change Δ7 may require the integration time of detection equipment to be increased to counter the reduced signal strength. Consequently, the screening rate will depend upon this trade off between temperature change Δ7 and integration time. For the photothermal detection apparatus 50 this is detailed below in table !

Table 1

These values also depend on the diameter of nanoparticles to be detected. The values in table 1 are with respect to gold nanoparticles of 20nm diameter. Larger particles provide stronger signals and thus are easier to detect. Therefore, when larger particles are used the modulated heating beam MHB can be used at lower power values.

In summary, the photothermal effect can be used to optically detect metal, typically gold nanoparticles freely diffusing in liquid medium. By combining photothermal microscopy with a heterodyne detection scheme single-particle sensitivity can be achieved in enlarged detection volumes for particle sizes down to a diameter of 20 nm. In back- scattering mode, the photothermal signal shows fast fluctuations due to diffusion of the particles across an interference pattern formed due to the backscattered and reference fields in the detection volume, which introduce a short component in the correlation function. This additional component is eliminated in forward-scattering mode. In the forward configuration described with reference to Figure 1 , interference effects are reduced and the signal intensity increases, thereby allowing smaller-sized particles to be detected, even for relatively large detection volumes. Use of metal nanoparticles as labels for biological systems is desirable due to the absence of photobleaching and blinking effects. This is important for screening applications with single-object sensitivity and high selectivity requirements, where multi-stage sorting steps will be necessary.

A skilled person will appreciate that variations of the disclosed arrangements are possible without departing from the invention. For example, although gold nanoparticles are used as the labels in the above descriptions, all types of inert metal nanoparticles, as well as semiconductor quantum dots, can be used to perform simultaneous screening of different targets, e.g. gold and silver nanoparticles. Also, although the screening applications with single-object sensitivity are described above in isolation, it will be appreciated that the screening could be part of another larger process, for example, involving multi-stage sorting steps. Accordingly the above description of the specific embodiment is made by way of example only and not for the purposes of limitation. It will be clear to the skilled person that minor modifications may be made without significant changes to the operation described.

Claims

1. A photothermal system for performing high throughput single particle detection, the system comprising: a first optical path for directing a probe laser beam and a heating laser beam towards a target or sample area; and a second optical path for directing the probe laser beam from the sample to a detector for detecting phase changes induced in the probe beam by the photothermal effect due to local heating of the sample by the heating beam, wherein the first optical path includes one or more optical elements arranged to define a detection volume of more than 10 femtolitres.

2. A system as claimed in claim 1 wherein the detection volume is more than 20 femtolitres.

3. A system as claimed in claim 2 wherein the detection volume is more than 30 femtolitres, ideally more than 40 femtolitres, for example 50 femtolitres.

4. A system as claimed in any of the preceding claims wherein the first optical path includes one or more optical elements arranged to have a numerical aperture of less than 1.

5. A system as claimed in claim 4 wherein the numerical aperture is less than 0.8, for example 0.5.

6. A system as claimed in any of the preceding claims wherein the first and second optical paths are non-coincident.

7. A system as claimed in any of the preceding claims wherein the detector uses a reference laser beam to detect phase changes in the probe laser beam after it has passed through the sample.

8. A system for performing high throughput single particle detection, the system comprising: a first optical path for directing a probe laser beam and a heating laser beam towards a target or sample area; and a second optical path for directing the probe laser beam from the sample to a detector for detecting phase changes induced in the probe beam by photothermal effect due to local heating of the sample by the heating beam, wherein the first optical path includes one or more optical elements arranged to have a numerical aperture of less than 1.

9. A system as claimed in claim 8 wherein the numerical aperture is less than 0.8, for example 0.5.

10. A system for performing high throughput single particle detection, the system comprising: a first optical path for directing a probe laser beam and a heating laser beam towards a target or sample area; and a second optical path for directing the probe laser beam from the sample to a detector for detecting phase changes induced in the probe beam by photothermal effect due to local heating of the sample by the heating beam, wherein the first and second optical paths are non-coincident.

11. A screening process comprising: screening a target substance for active particles to which labels are attached, for example metal nanoparticles, using the system of any of the preceding claims, and isolating particles when a positive detection result is obtained.

12. A process as claimed in claim 11 wherein when the target substance is a liquid or solution, the process further comprises separating a sample volume of the liquid when a positive detection result is obtained; collecting all separated sample volumes with positive detection results; diluting the collected sample volumes and re-screening the collected sample volumes in order to enrich the solution with active particles.