Relationship of Temperature Rise to Incident Laser Power in Optical Traps

Lauren I. Rugani

Optical trapping often is used to evaluate nanometer-size systems by introducing up to hundreds of milliwatts of laser power into the focus. The resulting power intensity can raise local temperatures and induce unwanted physical, chemical or biological reactions in the system of interest.

Syoji Ito and Hiroshi Miyasaka, together with a team at Osaka University in Japan, have applied fluorescence correlation spectroscopy to investigating the relationship between temperature rise and incident laser power under optical trapping conditions.

Using this setup, researchers performed fluorescence correlation spectroscopy to measure the relationship between incident laser power and increasing temperature in an optical trap.

The setup included a continuous-wave Ar⁺ laser at 488 nm from Lasos Lasertechnik GmbH that excited samples of fluorescent probes Rhodamine 6G and Rhodamine 123 in separate solutions of water, ethanol and ethylene glycol. A Spectra-Physics Nd:YVO₄ laser at 1064 nm supplied near-infrared light for the trap. It was also focused onto the sample. The researchers adjusted focus points from each laser to be 30 μm above the bottom of the glass culture dish, and an avalanche photodiode from PerkinElmer detected emitted photons.

With fluorescence correlation spectroscopy, the team measured the average value of the translational diffusion velocity of fluorescent molecules passing through the confocal area to determine the average temperature inside the sample volume. Measuring the transmittance of the NIR light through varying cylindrical sample volumes helped determine the extinction coefficients for each solvent: 14.5 m^–1 for water, 11.2 m^–1 for ethanol and 19.9 m^–1 for ethylene glycol.

When mapped as a function of path length, the transmitting laser power decreased exponentially with increased path length. The scientists suggest, however, that the experimentally obtained extinction coefficients indicate only linear absorption of NIR irradiation by the solvent, and, thus, the nonlinear process can be ignored.

They derived an autocorrelation function of the fluorescence intensity fluctuation at the confocal volume. The analysis provided information on molecular motion, the translational diffusion coefficient and the diffusion time, which relates to solution viscosity and temperature. Independent viscosity measurements with an Ostwald viscometer, along with the value obtained by fluorescence correlation spectroscopy, allowed the researchers to estimate the temperature, provided that the diameter of the probe molecule remained constant.

Plotting the fluorescence correlation function against time showed that the calculated autocorrelation curves agreed with experimental data and decayed faster with increased NIR laser power. The researchers attribute this change to accelerated molecular motion caused by an increase in temperature and not by the optical force. Analysis of the curves combined with the previous temperature-viscosity dependencies shows a linear increase in temperature at the focal point with incident laser power up to 240 mW.

The values of ΔT/ΔP were found to be 23 ±1 K/W for water, 49 ±7 K/W for ethanol and 62 ±6 K/W for ethylene glycol. This proportionality increases as a function of the extinction coefficient over the thermal conductivity of the solvents, with a slight temperature deviation at the focus of the NIR laser light. The method allows temperature measurements in areas as small as 350 nm as a result of the high sensitivity and low concentration of fluorescent probes.

The researchers propose that the technique has potential for monitoring temperature in microfluidic systems or for measuring local temperatures of cell activities.

Journal of Physical Chemistry B, March 8, 2007, pp. 2365-2371.