Method for real-time monitoring of paramagnetic reactions using spin relaxometry with fluorescent nanodiamonds

Abstract

Spin relaxometry using fluorescent nanodiamonds (FNDs) has been applied successfully to sense numerous paramagnetic target molecules such as free radicals and metalloproteins. However, despite their high sensitivity, T1 spin relaxation measurements are often hampered by their slow acquisition speed. Here, we demonstrate a method that allows for real-time monitoring of paramagnetic chemical reactions. We demonstrate T1 spin relaxometry from thousands of FNDs using an optimised cuvette-based system integrating an avalanche photodiode operated in linear mode, and a fast, fieldprogrammable gate array (FPGA) for data collation. We demonstrate chemical monitoring of the reduction of Cu(II) to Cu(I) ions in-solution with a 15 second integration using an optimised T1 sensing protocol. Our method achieves more than two orders of magnitude speed up with an order of magnitude reduction in cost when compared with traditional techniques. With further technical improvements, we believe this in-solution method could be extended to sense the sub-second chemical kinetics of paramagnetic molecules in solution.

Figures (3)

• Figure 1: a) In-solution measurements are conducted on dispersed FNDs in a standard cuvette using a green (532 nm) excitation laser which generates red fluorescence from the NV defects b) Schematic of the energy levels of the NV defect consisting of ground($^3A_2$) and excited state triplets($^3E$) and two intermediate singlet states ($^1A_1$,$^1E$).After excitation with green (532 nm) the system can decay back to the ground state radiatively (637 nm) or non-radiatively through the intermediate singlet states. c) A schematic of a the laser pulse sequence, resulting fluorescence, and transmitted data. d) A typical $T_1$ relaxation curve resulting from the processed data acquired from the pulse sequence in c. Plot compares the baseline measurement of FND dispersed in MilliQ (blue trace), and the same sample after addition Cu(II) ions (10 µ M) (orange trace). e) Rendering of the 3D printed cuvette holder with integrated PD module connected to FPGA development board which acquires the voltage signal from the PD module.
• Figure 2: a) Comparison of the $T_1$ curves obtained using the previous SPC and the new PD-based system. b) Comparison of the fit error on the $T_1$ time for the SPAD and PD systems. c) The $T_1$ relaxation time measured for different concentrations of 50 and 70 nm FNDs using the PD-based system. d) Intra- and inter-sample variation using 50 and 70 nm FNDs on the PD-based system. Five aliquots from the 50 and 70 nm samples were taken and then each measured five times. All measurements are shown.
• Figure 3: a) Reaction schematic of the reduction of paramagnetic Cu(II) by ascorbic acid to magnetically silent Cu(I). b) Diagram of the sample preparation and chemical reaction procedure. 50 nm FND (100 µ g/mL) were prepared in MES buffer (25 mM, pH 6) and degassed with nitrogen before sealing with a layer of mineral oil. An aliquot of Cu(II) solution (1 µ L,4 mM) was injected through the oil layer. To start the chemical reaction an excess of ascorbic acid (1 µ L, 500 mM) was injected through the oil barrier. c) The fitted $T_1$ time from the full $T_1$ curve at two minute intervals at each step of the reaction protocol shown in b) (blue trace). The error bars on the points represent the fit error on the $T_1$ parameter from Equation (1). Orange shaded region shows the change in PL signal and associated variance over the course of the reaction protocol reprocessed from the full trace data using a 3 point sequence. Vertical dotted lines represent each step in the protocol.