Phase-transition thresholds and vaporization phenomena for ultrasound phase-change nanoemulsions assessed via high-speed optical microscopy

Abstract

Ultrasonically activated phase-change contrast agents (PCCAs) based on perfluorocarbon droplets have been proposed for a variety of therapeutic and diagnostic clinical applications. When generated at the nanoscale, droplets may be small enough to exit the vascular space and then be induced to vaporize with high spatial and temporal specificity by externally-applied ultrasound. The use of acoustical techniques for optimizing ultrasound parameters for given applications can be a significant challenge for nanoscale PCCAs due to the contributions of larger outlier droplets. Similarly, optical techniques can be a challenge due to the sub-micron size of nanodroplet agents and resolution limits of optical microscopy. In this study, an optical method for determining activation thresholds of nanoscale emulsions based on the in vitro distribution of bubbles resulting from vaporization of PCCAs after single, short (<10 cycles) ultrasound pulses is evaluated. Through ultra-high-speed microscopy it is shown that the bubbles produced early in the pulse from vaporized droplets are strongly affected by subsequent cycles of the vaporization pulse, and these effects increase with pulse length. Results show that decafluorobutane nanoemulsions with peak diameters on the order of 200 nm can be optimally vaporized with short pulses using pressures amenable to clinical diagnostic ultrasound machines.

Figure 1

Dynamic light scattering measurements of DFB nanoemulsions produced by microbubble condensation. Lines show the result of sizing averaged for 3 repeated measurements per sample over 3 samples (9 total measurements), in both a number-weighted and intensity-weighted format. When number-weighted, the mode diameter occurs at 164 nm, with a mean droplet diameter of 192 ± 85 nm; when intensity-weighted, mode size occurs at 190 nm with mean size of 297 ± 323 nm. Note: Malvern Nano ZS set to ‘Multiple Narrow Modes’ for high-resolution analysis.

Figure 2

Experimental setup. A piston ultrasound transducer is manually triggered to deliver a short pulse to the portion of the microcellulose tube resting in the optical plane. Prior to droplet vaporization, large droplets can be visualized near the bottom of the tube with no significant presence of bubbles near the top of the tube. After vaporization is induced, the focal plane is adjusted to the top of the microcellulose tube to capture images of bubbles produced. Axes indicate the elevational, lateral, and axial dimensions of the incident ultrasound pulse.

Figure 3

Ultra-high-speed microscopy (20 million frames per second) of bubbles produced during a 1 MHz, 20-cycle vaporization pulse of 1.45 MPa rarefactional pressure. Many bubbles are observed to emerge through the first rarefactional phase (a–d), but as the transition through the second compression phase occurs (e–g) the bubbles are observed to compress and disappear from view. Upon the second rarefactional phase (h,i), many of the bubbles re-emerge (denoted by white arrows), but those indicated by black arrows in (d) do not re-appear. During the third compression phase (j – i), the bubbles are observed to compress again. Scale bars indicate 5 µm. Note: This image comprises 12 selected frames from a longer source video. See Supplemental Data for source video.

Figure 4

Ultra-high-speed microscopy (1 million frames per second) of bubbles produced during a 1 MHz, 20-cycle vaporization pulse of 1.45 MPa rarefactional pressure. Droplets are observed to vaporize in the first rarefactional phases and fully expand within 1–2 µs (a–d). Microbubble phenomena of fusion (b–d, white arrows) and radiation force (d–i) can be seen, indicating that large bubbles may combine to form much larger bubbles under the influence of long vaporization pulses. Scale bars indicate 5 µm. Note: This image comprises 9 selected frames from a longer source video. See Supplemental Data for source video.

Figure 5

Representative vaporization pulses at (a) 1, (b) 5.5, and (c) 8 MHz when driven with a 2-cycle sinusoid. Low transducer bandwidth for the high-power piston transducers resulted in additional ringing behavior such that droplets were exposed to more than just two compression/rarefaction phases. Non-linear propagation effects are clearly visible at these pressures.

Figure 6

Examples of bubble distributions generated as a function of ultrasound output at 5.5 MHz. At rarefactional pressures near (a) 2 MPa, few bubbles on the order of 1 µm in diameter were present, while increasing the pressure to approximately (b) 3 MPa and (c) 4.25 MPa increased the proportion of small bubbles present. Scale bar represents 5 µm.

Figure 7

Histogram of sparse bubbles present in 3 droplet emulsion samples (N=610) without the use of a vaporization pulse as a result of spontaneous thermal vaporization. Without ultrasonic vaporization, the small number of bubbles present had a mean size of 8.6 ± 3.3 µm, with a mode of approximately 3 µm.

Figure 8

Changes in distribution statistics as a function of frequency and rarefactional pressure averaged for 3 separate samples. Mean diameter (a) decreased in a generally linear fashion as rarefactional pressure increased, and all samples settled at a similar mean diameter at the highest ultrasound pressure that could be delivered. Mode diameter (b) fluctuated highly at the lowest pressures used for each frequency, but once pressures increased past a threshold, the mode settled to the lowest resolvable bin size – indicating bubbles primarily resulted from activation of the modal peak. The rarefactional pressure required to achieve a similar mode size was seen to increase with frequency.

Figure 9

Histogram of all bubble sizings taken at 5.5 MHz (N=11,654) as a function of rarefactional pressure (bin size 0.5 µm). At the lowest pressures tested, the distribution peak shifts to much higher than the controls with no ultrasound (figure 7). As pressure increases, the number of small bubbles increases in proportion until the smallest bin size overtakes as the peak in the distribution. Continued increase in the pressure amplifies the proportion of these small bubbles relative to other bubbles present in the distribution.

Figure 10

Measures of distribution skew (Pearson’s first skewness coefficient) averaged for all samples at each frequency. At the lowest pressures used, skew was near zero, but increased with pressure until a transitional pressure – past which skew remained constant or decreased slightly. The transitional pressure (indicating bubbles primarily resulted from nanoscale droplets in the modal peak) appears to increase with frequency.