Microfluidic oxygen tolerability screening of nanocarriers for triplet fusion photon upconversion

The full potential of triplet fusion photon upconversion (TF-UC) of providing high-energy photons locally with low-energy excitation is limited in biomedicine and life sciences by its oxygen sensitivity. This hampers the applicability of TF-UC systems in sensors, imaging, optogenetics and drug release. Despite the advances in improving the oxygen tolerability of TF-UC systems, the evaluation of oxygen tolerability is based on comparing the performance at completely deoxygenated (0% oxygen) and ambient (20–21%) conditions, leaving the physiological oxygen levels (0.3–13.5%) neglected. This oversight is not deliberate and is only the result of the lack of simple and predictable methods to obtain and maintain these physiological oxygen levels in an optical setup. Herein, we demonstrate the use of microfluidic chips made of oxygen depleting materials to study the oxygen tolerability of four different micellar nanocarriers made of FDA-approved materials with various oxygen scavenging capabilities by screening their TF-UC performance over physiological oxygen levels. All nanocarriers were capable of efficient TF-UC even in ambient conditions. However, utilizing oxygen scavengers in the oil phase of the nanocarrier improves the oxygen tolerability considerably. For example, at the mean tumour oxygen level (1.4%), nanocarriers made of surfactants and oil phase both capable of oxygen scavenging retained remarkably 80% of their TF-UC emission. This microfluidic concept enables faster, simpler and more realistic evaluation of, not only TF-UC, but any micro or nanoscale oxygen-sensitive system and facilitates their development and implementation in biomedical and life science applications.


1
Size and polydispersity index of the nanocarriers The size and polydispersity index of the nanocarriers was determined with dynamic laser scattering at 22 ℃ in deionized water. The size (diameter) and polydispersity indexes of the nanocarriers are shown in Table S1. On-chip oxygen levels The oxygen levels inside the microfluidic chip were determined by using an optical oxygen sensor (Piccolo2, Pyro Science, Aachen, Germany) with nanoprobes (OXNANO, Pyro Science, Aachen, Germany) dispersed in water. The oxygen levels at the measurement spot at the flow rates used in the study are shown in Table S2.

Upconversion measurements
The schemes of the upconversion measurements from cuvette and from microfluidic chips are shown in the Figures S1 and S2.

Determination of upconversion quantum yields
The upconversion quantum yields (QY) at 0 % oxygen (by using 30 mM of Na2SO3) were determined by integrating the emission spectra of the nanocarriers from 400 to 520 nm and using dilute (maximum absorbance less than 0.1) Rhodamine 6G in ethanol as a reference (95 % fluorescence quantum yield 1 , emission spectrum integrated from 525 to 800 nm). The quantum yields were then calculated using the following equation 2 : where Φ std is the quantum yield of the reference, A is the absorbance at the excitation wavelength, I is the integrated emission intensity and η is the refractive index of the medium 3 (1.33 for water and 1.36 for ethanol). Note that the quantum yields were determined without the multiplication factor of 2.
The 532 nm excited emission spectra of the nanocarriers are shown in Figure S3.

Figure S3
Emission spectra of the nanocarriers at anoxic conditions under 3100 mW/cm 2 excitation, which yielded the maximum quantum yield of upconversion. Scattering from the 532 nm excitation is excluded for clarity.

Figure S4
Quantum yield of upconversion under varied excitation power density of each nanocarrier system.
Upconversion performance at ambient conditions was determined by comapring the emission spectra measured at ambient conditions with the emission spectra obtained at anoxic conditions. The emission spectra (at anoxic and ambient conditions) obtained at 1000 mW/cm 2 and 260 mW/cm 2 excitation are shown in Figures S4 and S5.

Stability of nanocarriers under excitation in ambient conditions
The UC emission was monitored over 1 hour period (see Fig. S4) in ambient conditions (the cuvette was uncapped) under 1000 mW/cm 2 excitation. The results were normalized by the UC emission yielded by the nanocarriers at anoxic conditions. To study the photobleaching of the samples, absorption spectra were measured before and after 1 hour of excitation (see Fig. S5, shown for T80 + EO and T80 + CEL, diluted 10 times). No changes in the PtOEP Q band absorbance were noticed with any samples and thus the slight decrease in absorbance between 325 and 425 nm in the samples was attributed to the photobleaching or photooxidation of DPA. CEL + EO and T80 + EO showed less photobleaching (approximately 4 %) than CEL + M840 and T80 + M840 (approximately 10 %). Figure S7. UC emission of the nanocarriers excited at 1000 mW/cm 2 and monitored for 1 hour. Figure S8. Absorption spectra of T80 + EO and T80 + M840 nanocarriers before and after 1 hour excitation at 1000 mW/cm 2 . Photobleaching of DPA is visible in the slight decrease of absorbance between 325 and 425 nm.

Upconversion emission measurements from the microfluidic chip
The upconversion measurements from the microfluidic chip were performed with an excitation power density of 140 mW/cm 2 and the upconverted emission was collected through a 520 nm short-pass filter. The measurement range of the spectrometer was 400-800 nm and thus the emission spectra of 9,10-diphenylanthracene was slightly cut. No changes in the shape of the absorption and emission spectra between the nanocarriers were observed. The emission spectra from T80 + EO nanocarriers under 532 nm excitation is shown in Figure S1. Upconversion emission intensities at varied oxygen levels normalized by quantum yields The upconversion intensities measured from the microfluidic chip were normalized by the QY of each nanocarrier at 140 mW/cm 2 . These results are shown in Figure S2.