Skip to content

Nanodiamond Sensing of the Transmetalation Kinetics of Gd-DTPA in Individual Levitated Microdroplets

FieldValue
JournalThe Journal of Physical Chemistry B
Published2026-06-16
AuthorsEmily K. Brown, Zachary R. Jones, Adrisha Sarkar, Brandon J. Wallace, Ashok Ajoy
DOI10.1021/acs.jpcb.6c01528

The ODMR-derived rate constant for Gd-DTPA transmetalation was 4.8-fold lower in individual levitated microdroplets than in bulk solution at 0.05 M HCl. Without added HCl, bulk solution produced a 25% ODMR response, while droplets showed no change. The droplets averaged 18.6 µm in radius.

Transmetalation releases paramagnetic Gd³⁺ as Zn²⁺ captures DTPA. NV centers in carboxy-terminated nanodiamonds detect free Gd³⁺ through magnetic-noise pathways that reduce ODMR contrast. The readout remains continuous and does not require extracting the analyte from the droplet.

Within the examined concentration range, efficient detection requires free Gd³⁺ to adsorb to the negatively charged nanodiamond surface. The sensor did not efficiently detect Gd³⁺ while it remained bound to DTPA.

Zn²⁺ was selected because it is diamagnetic and therefore invisible to the sensors. Adding 48 mM zinc acetate at pH 6.6 changed the ODMR response by 2.6 ± 3.3%. Adding 1.5 mM Gd-DTPA at pH 6.4 changed it by -3.1 ± 2.8%. These controls associate the kinetic signal with released Gd³⁺.

All nanodiamonds were carboxyl terminated and contained about 2 ppm NV centers. Droplet experiments used approximately 100 nm particles, while bulk experiments used 140 nm particles. Every solution was a 55:45 mixture of HPLC-grade water and glycerol. Its water activity of 0.8 prevented droplet evaporation during measurements.

One charged droplet containing Gd-DTPA and nanodiamonds collided with an oppositely charged droplet containing zinc acetate and HCl. The mean reactant concentrations after collision were 1.5 mM Gd-DTPA and 48 mM zinc acetate.

The estimated nanodiamond concentration was 3 × 10¹¹ mL⁻¹. Settling in the dispenser could raise that concentration by as much as 30 times. Each data point averaged at least three droplet collisions.

Trap humidity was maintained at 80 ± 3.5% with a combined nitrogen flow of 0.1 SLM. A piezoelectric dispenser generated the droplets. Mie scattering measured their radius with ±0.5 µm accuracy, and the radius standard deviation remained below 0.3 µm.

Microwave excitation at 2.865 GHz was amplitude modulated at 1 kHz. A 532 nm laser excited the NV⁻ centers, and fluorescence above 650 nm was collected by a photomultiplier tube.

Parallel NMR measurements benchmarked the nanodiamond rate constants. The NMR experiments used a 500 MHz spectrometer, a 5 mm BBO probe, and a temperature of 293 K. Measurements of ¹H T1 relaxation were recorded before mixing and about every minute after mixing.

Both methods used 1.5 mM Gd-DTPA and 48 mM zinc acetate. NMR measurements also confirmed that nanodiamond addition did not alter the observed reaction kinetics.

The reaction took about 15 to 30 minutes over the pH range of 6.2 to 4.3. At pH 3.9, NMR showed completion in under 15 minutes. The nanodiamond kinetics gave the same overall timescale, although no appreciable ODMR change occurred during 10 minutes at pH 2.7.

The rate constants from NMR and nanodiamond ODMR agreed closely, particularly above pH 4. They also agreed with the model developed by Sarka and colleagues. That model had been tested at 5 × 10⁻⁴ M Gd-DTPA, 0.003 to 0.026 M Zn²⁺, and pH 4.62 to 5.21.

Above pH 4, Zn²⁺ attack was the main reaction pathway. Below about pH 4, protonation increased the observed rate. The model underestimated both measurements at pH 3.9. Assigning 1 × 10³ M⁻¹ s⁻¹ to the acid-assisted metal pathway improved agreement at that pH.

Quantifying free Gd³⁺ requires condition-specific calibration curves. These curves were nonlinear and showed Langmuir-type behavior because Gd³⁺ must adsorb to the nanodiamond surface. Acetate complexation shifted the curves toward higher apparent Gd³⁺ concentrations.

The calibration was strongly pH dependent. At 1 mM Gd³⁺, the ODMR response decreased from 70% at pH 6.2 to 31% at pH 2.7. Protonation of surface carboxyl groups reduces the adsorption sites available for Gd³⁺.

Gd³⁺ adsorption was faster than the monitored reaction. The ODMR signal from added GdCl₃ appeared within 30 seconds of addition, before the first kinetic measurement.

The nanodiamond measurements predicted a maximum released concentration of 10 µM at pH 6.2. NMR and Visual MINTEQ speciation instead indicated that nearly all 1.33 to 1.45 mM Gd-DTPA dissociated under the tested conditions.

The authors attributed the concentration disagreement mainly to pH-dependent suppression of sensor sensitivity. Matrix effects from reaction components and nanodiamond aggregation were also possible. The unresolved discrepancy limits absolute concentration measurement, but the rate constants extracted from both methods remained similar.

Droplet pH was not measured directly. Small volume can cause deviations of several pH units from the bulk solution, and the authors estimated droplet composition with a model that accounts for acetic acid and HCl partitioning.

HCl in bulkMeasured bulk pHHCl in dropletEstimated droplet pH
0 M6.20 Mabout 6.2 to 7
0.05 M4.70.05 Mabout 6 to 7
0.07 M4.30.14 Mabout 1.3 to 1.8
0.09 M3.90.24 Mabout 0.8 to 1.5
0.12 M2.7

The model predicts that all acetate evaporates above 0.05 M HCl. Below that concentration, the estimated pH rises above the pKa of acetic acid, which slows its evaporation.

Assuming acetate evaporation alone gives predicted pH values of 1.3 and 0.8 at 0.14 M and 0.24 M HCl. ODMR responses above 10% at those concentrations suggest that HCl evaporation also occurred.

A reported HCl evaporation rate of 1 × 10⁻³ s⁻¹ in approximately 4 µm droplets at 80% relative humidity could raise the pH over 30 minutes. The predicted values reach 2.1 and 1.6 for initial HCl concentrations of 0.14 M and 0.24 M.

The uncertain droplet pH led the authors to compare results by HCl concentration rather than pH. A 0.24 M HCl droplet produced a 12% ODMR change. Rate constants fitted to ODMR response were 1.5 times larger than constants fitted to Gd³⁺ concentration because the calibration curves were nonlinear. The authors assumed comparable nonlinearities for the relative bulk and droplet comparison.

Nanodiamond ODMR can report relative reaction rates in bulk solution and individual microdroplets. It does not provide reliable free Gd³⁺ concentration without calibration for the reaction pH, acetate concentration, and surface conditions.

The controls separate two effects. High pH preserves sensor sensitivity but suppresses transmetalation. Low pH increases proton-assisted release while reducing the maximum ODMR response. Either effect can resemble a change in reaction rate.

The strongest kinetic result is the consistent slowing in droplets across the tested HCl range. A direct bulk comparison is valid for relative rates because both datasets use the same reaction and ODMR readout. It is not valid for comparing absolute free Gd³⁺ concentrations.

The cause of the droplet deceleration remains unresolved. The authors list interfacial electric fields, surface partitioning, partial solvation, and radical formation as mechanisms proposed for microdroplet reactivity. Some of these processes can slow a reaction under particular conditions.

Preferential partitioning of Gd-DTPA to the droplet interface could reduce its availability in the droplet core. The authors state that this remains a possible explanation. A systematic measurement of rate constant against droplet radius could test for a surface contribution. A 1/radius dependence would support a surface-mediated component.

The paper was posted on 1 April 2026 as a preprint that had not been peer reviewed, and the authors note that the data may be preliminary. For sensor development, the results support NV-center nanodiamond as an in situ kinetic reporter when relative rates are sufficient. They do not support using ODMR amplitude alone to determine free Gd³⁺ concentration.

Any microdroplet assay using this approach should report HCl concentration, estimate droplet pH, and include a matched bulk control. Surface termination and pH range also require calibration before reaction rates can be compared across devices.

View Original Abstract

Directly measuring kinetics within microcompartments is critical for determining the distinct mechanisms that govern chemistry in confined environments. Here, we deploy nanodiamond (ND) particles hosting nitrogen-vacancy (NV) centers as quantum sensors to track the transmetalation of gadolinium diethylenetriaminepentaacetic acid (Gd-DTPA) in bulk solution and within single levitated, charged microdroplets by optically detected magnetic resonance (ODMR). The appearance of paramagnetic Gd + during the reaction produces changes in the continuously acquired ODMR signature, furnishing an in situ optical measurement of kinetics in microscale volumes that are otherwise difficult to interrogate directly. While quantification of [Gd + ] remains challenging due to the high sensitivity of ODMR to pH and ligand concentration, we find that NDs accurately report on observed rate constants for the reaction over a broad pH range by comparison with NMR measurements. A direct comparison between droplet and bulk ND measurements also allows for a reliable kinetic comparison between the two environments. The transmetalation rate in droplets, with an average radius of 18.6 μm, proceeds significantly slower than in a bulk solution, with up to a 4-fold reduction in the observed rate constant. Additionally, in the absence of added acid, the transmetalation reaction proceeds efficiently in bulk, yet is completely suppressed in droplets, suggesting that microconfinement can modulate reactivity. These findings highlight both the strengths and limitations of ND-based ODMR sensing for probing reaction kinetics in microscale environments.