Skip to content

A Compact Incubation Platform for Long‐Term Cultivation of Biological Samples for Nitrogen‐Vacancy Center Widefield Microscopy

FieldValue
JournalAdvanced Materials Technologies
Published2026-03-25
AuthorsAndre Pointner, Daniela Thalheim, Sarah Belasi, Lukas Heinen, LR Carnell
DOI10.1002/admt.70949
Citations1

A purpose-built chamber maintained HT29 colorectal cancer cells through 90 hours of continuous incubation. It controlled temperature, CO2, and humidity while preserving the total internal reflection geometry required for nitrogen-vacancy center widefield magnetometry. A shorter 48-hour test on the functionalized diamond showed cell viability, proliferation, and magnetic dipole patterns from superparamagnetic iron oxide nanoparticles. The 90-hour and 48-hour tests used different surfaces, so the results set separate limits for cover-slide and direct-diamond cultivation.

The incubation problem created by TIR imaging

Section titled “The incubation problem created by TIR imaging”

NV widefield magnetometry excites NV centers through total internal reflection rather than objective-based illumination. A 532 nm laser entered the side of a polished glass cube and reached the lower diamond surface at 36° to 39°, depending on the field of view. The beam reflected at the upper diamond surface and away from the biological sample.

This geometry limits excitation light reaching the sample, but it does not match commercial stage-top or whole-microscope incubators designed for conventional inverted microscopes. The authors therefore enclosed the glass cube and diamond in a separate chamber. Electronics, optics, stages, and bias magnets remained outside the humid atmosphere.

The sensing specimen was [100]-oriented bulk diamond with a 1 µm thick layer containing 1 ppm NV centers. It was fabricated by Quantum Diamonds and mounted on the polished glass cube. An enameled copper wire looped around the diamond at approximately 36° to apply a uniform oscillating field perpendicular to the selected NV axis. Two samarium cobalt ring magnets, separated by 70 mm, generated a static field of up to 30 mT along that axis and magnetized superparamagnetic iron oxide nanoparticle labels.

A long-working-distance objective collected red brightfield light and NV fluorescence. The full brightfield field was 652 µm × 652 µm, while magnetic-field analysis used a 318.5 µm × 318.5 µm field at 1 kW cm⁻² laser intensity. The camera signal was binned 2 × 2 to improve the signal-to-noise ratio.

A dual-tone microwave signal drove both ¹⁵N hyperfine transitions with a 1.5 MHz intermediate frequency and a local oscillator suppression threshold of −30 dBm. Interleaved images with and without microwave excitation produced optically detected magnetic resonance contrast. The authors evaluated the center of three resonances at each pixel, converted the electron Zeeman shift to a local magnetic field, and applied rolling-ball background subtraction.

The chamber used an acrylic cylinder bonded to the glass cube with NOA61 optical adhesive. A rubber gasket and an upper chamber formed a closed atmosphere. The custom geometry used BioMed White Resin and was printed by stereolithography.

The glass cube was held by a heated aluminum assembly, while a separate heated lid sat 2°C to 3°C above the incubation temperature to limit condensation. An Arduino Nano ran PID control with thick-film resistors and NTC thermistors. The setpoint was about 34°C, slightly below 37°C, to avoid significant heating above 37°C if thermistor beta values shifted.

Calibration used deionized water and a sensor placed on the diamond. The measured gradient between the aluminum heater and diamond surface was 3°C. This allowed feedback control without a sensor inside the chamber that could affect ODMR. The microwave drive produced no measurable heating in either deionized water or culture medium.

A gas mixer supplied premixed 5% CO2 and 95% air, which passed through a humidifier before entering the chamber. A water reservoir and an internal overflow maintained high humidity and returned supply-line condensate to the reservoir. Grooves at the top of the acrylic cylinder allowed limited gas exchange so CO2 could diffuse into the cultivation volume while direct air exchange was limited. The design therefore addressed evaporation and the associated pH change in the small medium volume.

Cell culture validation and surface adhesion

Section titled “Cell culture validation and surface adhesion”

The authors seeded 500 µl of medium containing approximately 40000 HT29 human colorectal cancer cells onto a Fibronectin-functionalized microscope cover slide. Periodic brightfield imaging continued for 90 hours. The cells attached and divided during the full interval, which the authors used to confirm viability and proliferation. This test validated environmental control in the chamber on a cover slide.

Cultivation directly on the diamond exposed a separate materials problem. Cells had difficulty attaching to bare diamond. Fibronectin and Poly-L-Lysine each improved adhesion during 48-hour continuous tests. Fibronectin provided a more natural extracellular matrix. Poly-L-Lysine produced faster initial attachment through its electrostatic action, so the authors selected it for subsequent direct-diamond experiments.

The surface coating was therefore part of the working biological interface, not a minor preparation step. Bare diamond did not provide the adhesion needed for the intended measurement.

Brightfield images at the start, middle, and end of the 48-hour direct-diamond test showed typical morphology, adhesion, and proliferation. After incubation, the NV magnetometer recorded dipole patterns from the labeled cells. The authors attributed the pattern deformations to morphology changes and proliferation, which altered the homogeneous distribution of magnetic labels across the cell surface. This measurement showed that ODMR imaging remained functional after cultivation on the diamond.

The reported magnetic-field image also exposed optical limits. Signal became noisy near the upper and lower image edges because excitation intensity decreased there. The Gaussian beam profile projected onto the diamond under the TIR angle, which elongated the effective field along the x-axis and made the field elliptical. The paper does not report a numerical magnetic-field strength, before-and-after ODMR contrast, or sensitivity change for the cultured cells.

The paper’s comparison with prior incubation devices is functional rather than numerical. Commercial stage-top and whole-microscope incubators were described as incompatible because they are designed for objective-based inverted imaging. The compact chamber instead controls the atmosphere around the sensing surface without changing the TIR measurement path.

The internal time points need separate interpretation. Cultivation reached 90 hours on the functionalized cover slide, while the direct-diamond culture and magnetometry test lasted 48 hours. The 90-hour result does not establish 90 hours of direct cultivation on the diamond. The 48-hour experiment provides the more direct test because it combined growth on the sensing surface with a subsequent magnetic-field image.

Several measurement limits remain. The article gives no quantitative fluctuation ranges for chamber temperature, CO2 concentration, or humidity. It also gives no numerical viability fraction. Viability was judged from adhesion, morphology, and proliferation observed in brightfield images.

The optional fluorescence path could support common viability dyes, but the paper does not present dye-based viability results. Likewise, the 90-hour sequence used brightfield monitoring, and the direct-diamond magnetic image was acquired after 48 hours rather than as a reported continuous ODMR time series.

The authors identify real-time CO2 monitoring with closed-loop gas control as a future improvement. They also propose perfusion to support cultivation beyond 90 hours with controlled nutrient supply and waste removal. A multi-well version would allow parallel samples under identical conditions. These proposals address the lack of gas feedback and the stated need for nutrient supply and waste removal.

The paper’s material requirement is single-crystal diamond for an NV sensing surface, with a [100] orientation and a 1 µm, 1 ppm NV layer. The 6CCVD diamond materials catalog includes MPCVD-grown single-crystal plates and wafers up to 125 mm, while the orientation, concentration, and layer specification remain items to confirm. Use the engineering support link to define custom mounting and surface requirements for the TIR geometry.

View Original Abstract

ABSTRACT Nitrogen‐vacancy (NV) centers in diamond provide a versatile quantum sensing platform for biological imaging through magnetic field detection, offering unlimited photostability and the ability to perform long‐term observations without photobleaching or phototoxicity. However, conventional stage‐top incubators are incompatible with the unique requirements for NV widefield magnetometry to study cellular dynamics. Here, we present a purpose‐built compact incubation platform that maintains precise environmental control of temperature, atmosphere, and humidity while accommodating the complex constraints of NV widefield microscopy. The system employs a 3D‐printed biocompatible chamber with integrated heating elements, temperature control, and humidified gas flow to create a stable physiological environment directly on the diamond sensing surface. We demonstrate sustained viability and proliferation of HT29 colorectal cancer cells over 90 hours of continuous incubation, with successful magnetic field imaging of immunomagnetically labeled cells after extended cultivation periods. This incubation platform enables long‐term cultivation and real‐time monitoring of biological samples on NV widefield magnetometry platforms, opening new possibilities for studying dynamic cellular processes using quantum sensing technologies.