Noncovalent force spectroscopy using wide-field optical and diamond-based magnetic imaging
Sean Lourette, Lykourgos Bougas, Metin Kayci, Shoujun Xu, Dmitry Budker
NNoncovalent force spectroscopy using wide-field optical anddiamond-based magnetic imaging
S. Lourette, a) L. Bougas, M. Kayci,
1, 3
S. Xu, and D. Budker
1, 5 Department of Physics, University of California, Berkeley, California 94720-7300, USA Institut für Physik, Johannes Gutenberg Universität-Mainz, 55128 Mainz, Germany DWI-Leibniz Institute for Interactive Materials, Aachen 52074, Germany Department of Chemistry, University of Houston, Houston, TX 77204, USA. Helmholtz Institute Mainz, Johannes Gutenberg University, 55099 Mainz, Germany (Dated: 28 August 2019)
A realization of the force-induced remnant magnetization spectroscopy (FIRMS) technique of specific biomolecularbinding is presented where detection is accomplished with wide-field optical and diamond-based magnetometry usingan ensemble of nitrogen-vacancy (NV) color centers. The technique may be adapted for massively parallel screeningof arrays of nanoscale samples.
I. INTRODUCTION
Detecting molecular targets with high specificity and sen-sitivity is of importance in various areas of research, rang-ing from drug-design applications to diagnostics and dis-ease monitoring. Existing techniques include those based onatomic force microscopy (AFM), surface plasmon resonance(SPR), optical tweezers, surface-enhanced Raman scatter-ing (SERS), acoustic force spectroscopy (AFS), and force-induced remnant magnetization spectroscopy (FIRMS). FIRMS measures molecular bond strengths through the useof gradually increasing forces, typically applied using a cen-trifuge, or recently ultrasound. This is done by detectingthe magnetic field produced by a collection of magnetic mi-crospheres (beads) that label specific molecular targets. Thebeads are magnetized and bound to a surface by the specificinteraction of interest, in the presence of a magnetic field. Themagnetic field is then removed and the applied force is in-creased, breaking each bond as the applied force exceeds itsrupture force. This results in an abrupt change in the observedmagnetic field from the collection of magnetic beads corre-sponding to the strength of the specific interaction.
FIRMS, through its high force resolution, can precisely dis-tinguish noncovalent intermolecular interactions through highforce resolution, for example those of DNA duplexes witha single basepair difference. FIRMS offers high force reso-lution that can precisely distinguish noncovalent intermolec-ular interactions, for example those of DNA duplexes witha single basepair difference.
However, FIRMS currentlyuses vapor-based atomic magnetometers for detection, withsensor sizes on the order of millimeters. Due to the sizemismatch, the atomic magnetometers cannot detect individ-ual magnetic beads that are used to label the biomolecules.A more optimal sensor would detect magnetic fields on themicron scale, matching the size of the beads, and could re-solve individual beads while detecting much larger magneticfields, measuring closer to the source. We set out to im-prove upon FIRMS by shifting the detection to magnetic mi- a) Electronic mail: [email protected] croscopy, with a goal of creating a compact and robust instru-ment capable of background-free, optical diffraction-limitedimaging and high-throughput quantitative characterization ofbond strength.In this work, a proof-of-principle experiment is realized,combining FIRMS with magnetic and optical microscopy inorder to detect biotin-streptavidin bonds, chosen for theirwell-characterized interaction.
Streptavidin-coated mag-netic microspheres are allowed to bind to biotinylated andcontrol (non-biotin) diamond surfaces, which are then shakenwith piezo-electric actuators. In this study we replace the va-por cell magnetometer with a planar ensemble of nitrogen-vacancy (NV) centers, thus boosting the detected signal bymany orders of magnitude, providing optical diffraction-limited single-microsphere resolution. The biotin and controlsurfaces are prepared directly on the diamond plates, in orderto minimize the distance between the microspheres and theNV centers and to create a compact and robust instrument.Both magnetic imaging and wide-field optical imagingtechniques are used to detect detachment events. Mag-netic imaging provides background-free detection of the po-sitions and orientations of the particles, and is available inthe absence of optical access, such as when using an opaquemedium. Wide-field optical imaging offers higher throughput,allowing for the tracking of moving particles. Together, thesetwo imaging techniques provide correlated detection of indi-vidual particles and form a powerful toolset for quantitativelycharacterizing binding behaviors.We observe a distinct and reproducible binding behaviorwhen using control and functionalized surfaces, though theobserved rupture forces are in a different force regime thantypically observed in similar environments. Individual mi-crobeads are magnetically imaged and subtle changes in an-gular orientation between force applications are resolved. Thetechnique can be extended to nanoscale samples, enablingmassively parallel spatially resolved measurements. a r X i v : . [ phy s i c s . i n s - d e t ] A ug II. METHODSA. Diamond preparation
The diamonds used in this study are electronic gradechemical-vapor-deposition (CVD) plates, produced by Ele-ment 6, roughly 100 µm in thickness. Each plate was im-planted with N ions of energies and doses listed in the tablebelow, with the goal of creating a ≈
100 nm uniform high-density layer of NVs. N was chosen for its narrower spec-trum on account of its nuclear spin of 1/2 rather than spin 1. TABLE I. N Implantation ParametersEnergy (keV) Dose (cm − )10 4 ·
20 6 ·
35 9 ·
60 1 . ·
100 2 . · After implantation, the plates were annealed in an ovenfor 2 hours at 800 ◦ C and 4 hours at 1100 ◦ C. To prepare thesurface of the diamond to bind with streptavidin-coated mi-croparticles, the plates are functionalized with an Alkene-EG-COOH crosslinker, which bonds to biotin to form a biotiny-lated surface. Instead of using bare diamond, control plateswere also prepared through functionalization with an Alkene-EG-COOH crosslinker, omitting the step of bonding to biotin.Diamond functionalization was performed for us by AdamasNanotechnologies. B. Imaging magnetometer
The operating principle of the magnetometer is as follows.The NV center is composed of a substitutional replacementof a pair of adjacent carbon atoms with a nitrogen atomand a vacancy (Fig. 1a). It has a triplet spin ground state, m s = {− , , } , where at sufficiently low external fields, the0 sublevel has lower energy due to the electron-electron inter-action (Fig. 1c). The +1 and -1 sublevels are sensitive to themagnetic field, with the g-factor of 2 . / G. The magneticfield strength can be calculated by measuring the frequency ofthe 0 → + → − An ODMR spectrum is obtained by illuminating the NV withgreen light and optically detecting a reduction in fluorescencewhen a resonant microwave field is present (Fig. 1b). Thisreduction in fluorescence occurs due to a difference in the in-tersystem crossing rates among the three spin sublevels. A schematic of the imaging magnetometer is shown inFig. 2a. A 1 watt 525 nm diode laser (Roithner LasertechnikNLD521000G) is used to illuminate the diamond after reflect-ing off of a long-pass dichroic mirror (550 nm cutoff, ThorlabsFEL0550) and passing through an objective (Olympus UP-lanFl 40x). The collimation of the beam is adjusted in order to increase the size of the illuminated region at the focal planeof the objective, effectively increasing the field of view. Thepower of the light illuminating the sample is typically a fewhundred milliwatts. The fluorescent light is collected with theobjective, passes through the dichroic mirror, and is focusedonto a complementary metal–oxide semiconductor (CMOS)sensor (Thorlabs DCC1240M) forming an image of the NVlayer. A magnetic field of about 50 G is applied with a per-manent magnet and aligned to one of the four NV axes. Us-ing a 3D translation stage, a loop of wire is positioned within0.2 mm of the surface of the diamond, in order to apply mi-crowaves to the NV centers.The imaging magnetometer operates using a continuouswave (cw) excitation scheme: the 525 nm laser excitation isapplied while the microwave frequency is swept through theresonance of the m s = { → + } transition for the NV cen-ters aligned along the applied magnetic field. ODMR is em-ployed, and the resonance frequency is fitted for each pixel onthe camera to form a magnetic image.To enhance the signal quality, a lock-in detection scheme isused, in which the microwave source is switched on and offfor alternating camera frames (Fig. 1d). This is achieved bysynchronizing the camera trigger to a digital trigger of halffrequency, toggling between high and low for alternating im-ages, that controls the microwave output with an RF switch.The frequency of the camera trigger is set to be as high asthe camera would allow ( ≈
200 fps for 80x60 pixels) to avoidoverexposure while maintaining the maximal exposure time.To reduce the memory requirements of data storage andprocessing, instead of saving the raw frames, blocks of 32pairs of images are summed to produce a single pair of im-ages that are saved for later analysis. During analysis, theseimage pairs are converted to brightness and contrast images,
MicrowavesCameraLaser time
Microwave Frequency F l uo r e sc en c e (a) (b) (c)(d) FIG. 1. Operating principle of NV-based magnetometry. (a) The NVcenter is composed of a substitutional replacement of a pair of ad-jacent carbon atoms with a nitrogen atom and a vacancy. (b) TheODMR technique can be used to measure magnetic field strengthby sweeping the frequency of an applied microwave field, and ob-serving a decrease in fluorescence that appears on resonance. (c) Aspin-preserving optical transition between ground and excited tripletspin states with phonon sidebands. (d) The pulse sequence for cwmagnetometry with lock-in detection consists of a camera triggeringat twice the frequency of microwave modulation to ensure that mi-crowaves are present in alternating images, while the laser remainsactive throughout the experiment. The images in (a) and (c) are par-tially based on Refs. 15 and 16.
CMOSCameraLong-passFilterDichroicMirror Photodiode50/50Beamsplitter Neutral-density FilterNV plane L A S E R DiamondObjectiveNV plane L A S E R DiamondObjective
Imaging MagnetometerInterferometera)b)
FIG. 2. Optical schematics for (a) imaging magnetometer and (b)interferometer. (a) Fluorescence from the layer of NV centers iscollected with an infinity-corrected objective and passed through adichroic mirror and a long-pass filter, before being imaged onto aCMOS sensor with a lens. The NV centers are excited with a beamalso passing through the objective, generated by a 525 nm diodelaser. The beam from the laser is made to be slightly expanding in or-der to increase the size of the illuminated region in the NV plane. Forwide-field imaging with white light, a collimated white-light sourceilluminates the diamond surface directly at a normal angle of inci-dence, to create a bright background. (b) The dichroic mirror is re-placed with a 50/50 beam splitter, and a mirror is added to form theother arm of the interferometer. A graduated neutral-density filter isinserted to balance the reflected intensity from each arm of the inter-ferometer. The signal is detected on a fast photodiode and recordedon an oscilloscope. where the correlated low-frequency noise cancels out in thecontrast images. This process produces contrast images thatare insensitive to optical excitation spatial mode and power,thermal and mechanical drift, as well as CMOS sensor inho-mogeneities in both offset (dark signal non-uniformity) andgain (photo response non-uniformity) for fixed-pattern noise(FPN).To produce a magnetic image, the microwave frequency isrepeatedly swept at a rate of ≈ . The fittedcentral frequency is converted to a magnetic field, subtractedfrom the background field, and combined to form a magnetic field image using the gyromagnetic ratio ≈ . / G. C. Microfluidics
In this study, a sealed micro-fludic chamber is used tohold the solution containing the magnetic microspheres. Thechamber is designed to be as light as possible to increase themaximum force that can be applied to the magnetic particlesin-situ. A 3-D rendering of the chamber is depicted in Fig. 3.The top and bottom layers of the chamber are formed by glasscoverslips, and the walls of the chamber are formed by animaging spacer (Secure-Seal TM Spacer from Thermo FischerScientific), a 150 µm silicone gasket that is sticky on bothsides. Before sealing the chamber, the diamond is attachedto the center of the bottom coverslip using crystal wax withthe NV side face up, and the chamber is filled with a solutionof undiluted Phosphate Buffered Salene (PBS) and magneticmicrobeads. The diamond is attached to the lower (rather thanthe upper) surface, in order to allow gravity to assist the par-ticles in finding and bonding to the functionalized diamondsurface.The top side of the chamber is attached to a mount-ing post via two piezoelectric actuator stacks (ThorlabsAE0203D08F), which attach to opposite corners of the coverglass, leaving the bottom side accessible for viewing with anobjective (Olympus UPlanFl 40x). The objective is respon-sible for both carrying in the green light to the NVs for ex-citation and carrying out the red fluorescence to the camerafor imaging. The wire used to couple microwaves to the NVcenters is positioned as close as possible to the cover glassdirectly above the illuminated spot on the diamond, withoutitself being illuminated with the laser light.The device used to drive the piezo actuators is ThorlabsBPC303, a 150 V, 1 A closed-loop piezo driver. This devicewas used to drive the two piezo actuators in parallel with si-nusoids of frequencies ranging from 5 to 10 kHz, operatingclose to the 1 A current limit of the device.In order to avoid significant heating from the piezoelec-tric driving signal, the drive was broken up into sinusoidalpulses of 5 ms duration, applied once per second with fixedfrequency and varying amplitude. The amplitude was incre-mented after every 60 pulses from 0 V p − p to 50 V p − p in stepsof 2 V p − p .The magnetic microparticles used in this study are ferro-magnetic streptavidin-coated microspheres (SVFM-20) fromSpherotech. The microspheres are composed of polystyrenecores and CrO coatings, with an average density of 1.8g / cm and diameters distributed with mean, median, and stan-dard deviation of 2.10 µm, 1.95 µm, and 0.43 µm, respectively.A particle with diameter of 2 µm and density of 1 . / cm has an effective weight (weigh minus the buoyant force) of3 . × − N in PBS. Based on the distribution of sizes, onewould expect 68% of the particles to have an effective weightbetween 1 . × − N and 5 . × − N. FIG. 3. 3-D rendering of the chamber. The chamber is composed of an imaging spacer sandwiched between two glass coverslips. Beforesealing the chamber, the diamond is attached to the lower coverslip with crystal wax and the chamber is filled with a solution containing themagnetic microbeads. Adhesive is used to attach a pair of piezoelectric actuators (red) to opposite corners of the chamber on one side and toa mounting plate on the other. An objective (blue), which is positioned below the chamber, delivers the excitation beam (green) to the NVcenters and images the upper surface of the diamond onto a CMOS sensor. To provide the necessary microwave field to the NV centers, acopper wire (magenta) is positioned within 0.2 mm of the top surface of the chamber.
D. Interferometer
The strength of the force applied with the piezoelectric ac-tuators can be varied by changing the drive voltage. A forceis applied to a particle by using its own inertia to pull on thebonds at the apex of its mechanical motion. As such, the ap-plied force should be proportional to the amplitude of the me-chanical oscillation of the system, which in turn we expect tobe linear in applied voltage. The amplitude of the oscillationsof the mechanical system was measured with an interferome-ter, allowing the linear response to be verified, and the linearcoefficient to be obtained. With the measured amplitude andthe known drive frequency, a force per unit mass, or g-force,can be calculated. This allows us to map the applied force as afunction of frequency and voltage applied with the piezoelec-tric actuators.A Michelson-Morley interferometer was setup by swappingthe dichroic mirror for a 50/50 beam splitter, adding a mirrorto form the second arm of the interferometer, and insertinga graduated neutral-density filter to balance the reflected in-tensity from each arm of the interferometer (Fig. 2b). Theoutput of the interferometer is focused onto a photodiode,whose signal is amplified with a current amplifier (SRS570),while maintaining a bandwidth of 200 kHz, and subsequently recorded with an oscilloscope.In an interferometer, the two interfering beams produce anelectric field and intensity at the output that can be written as E ( t ) = Re { E e i ω t + E e i ( ω t + φ ) } = [ E + E cos ( φ )] cos ( ω t ) + [ E sin ( φ )] sin ( ω t ) , I ( t ) = (cid:104) E ( t ) (cid:105) = [ E + E + E E cos ( φ )] (1)where E and E are the two interfering fields, ω is their angu-lar frequency, and φ is the phase difference between the twoarms of the interferometer. Furthermore, the phase difference φ can be broken into two parts: the ideally constant piezo-independent phase distance φ , and a term governed by themechanical response to the piezo drive φ = φ + π x λ cos ( π f t ) , (2)where x and f are the amplitude and frequency of the mo-tion of the system, and λ is the wavelength of the light. Afterremoving the DC component, the predicted detected oscillo-scope voltage signal, proportional to intensity, will be of the -12 -10 a M ea s u r ed b M ea s u r ed c -12 -10 d P r ed i c t ed e P r ed i c t ed f FIG. 4. Power spectral density (PSD) of experimental photodiode data (a,b,c) alongside PSD of simulated data (d,e,f) with parameters chosento yield matching spectra. The experimental data were collected using piezo drive of 60 volts peak-to-peak at frequencies of 5.33 kHz (a), 7.67kHz (b), and 10.00 kHz (c). In each figure, the frequency axis has been scaled by 1 / f , such that the drive frequency and its harmonics appearat integers, allowing the value for N to be read directly from the plot (dashed vertical line). The fitted value for N can then be used to obtain ameasurement of the amplitude of the mechanical motion. form V ( t ) = V cos ( φ + N cos ( π f t )) . (3)Here N , which we have defined to be π x λ , can be estimatedby counting the number of harmonics of f before roll-off inthe power spectral density (PSD) of V ( t ) . In Fig. 4, The PSDof sample interferometer data (a,b,c) is displayed alongsidethe PSD of Eq. (3) (d,e,f), with matching values for N and f . III. RESULTS
Using wide-field imaging with white light (microscope il-luminator), the positions of the particles were tracked as theforce was applied, and the time at which each particle de-tached from the surface was obtained, in order to create a plotof number of bound particles as a function of applied voltage(Fig. 5). Wide-field imaging with white light was preferredhere, as it allowed for the simultaneous monitoring of dozensof particles. Reattaching particles were not counted; only par-ticles that started attached to the surface were tracked. Par-ticles were tracked manually, assisted by software written tofacilitate this process. The results were reproducible when re-peated for the same chamber; however, when a chamber wasbroken and a new one was assembled, results often deviatedsignificantly.After obtaining the particle statistics from the wide-fieldimaging with white light, the setup was switched to the in-
Peak-to-peak Voltage (V) P a r t i c l e s ( no r m a li z ed t o ) Bond Strength Measurements of Different Chambers
Chamber 1 (C)Chamber 2 (C)Chamber 3 (B)Chamber 4 (B)
FIG. 5. Normalized number of particles detached under increasinglylarger voltages for various chambers. Particle positions were trackedunder wide-field illumination as the voltage applied to the piezoelec-tric actuators was linearly ramped. The fraction of particles that re-main bound to the surface is plotted against the drive voltage. Thenumber of particles tracked during each experiment is indicated bya number labeling its corresponding line on the plot. Nine experi-ments were performed spanning four different chambers. Chamberswith biotin surfaces are labelled with (B) and chambers with controlsurfaces are labelled with (C).
Drive Frequency (kHz) F o r c e pe r V o l t ( p N / V o l t ) Mechanical Response of Different Chambers
Chamber 1 (C)Chamber 3 (B)Chamber 4 (B)
FIG. 6. Amplitude of the force experienced by a microsphere with di-ameter 2 µm as a function of drive frequency for different chambers.The force amplitude (F) is calculated based on the sinusoidal am-plitude of the mechanical motion ( x ), drive frequency ( ω / π ), andmass ( m ) according to the equation F = m ω x . Smoothing lines areadded to highlight trends. Chambers with biotin surfaces are labelledwith (B) and chambers with control surfaces are labelled with (C). terferometer configuration and the amplitude of the motion ofthe chamber was measured. Using the interferometer, severalchambers were analyzed to determine the applied force forvarious voltage drives. In Fig. 6, the measured force experi-enced by a particle of diameter 2 µm per unit of drive volt-age is plotted against the drive frequency for different cham-bers. The plot depicts variation in resonance features betweenchambers, with the force at a given drive frequency and ampli-tude varying among chambers by anywhere from 5% to 40%,depending on the frequency.The strength of biotin-streptavidin bonds has been demon-strated to depend heavily on the loading rate, the rate atwhich the applied force is increased. It should be noted thatour force application produces an oscillating force with an am-plitude that is slowly ramped, rather than a constant force thatis slowly ramped as is prevalent in the literature. During anoscillating force, the bond is only under stress for a fractionof each oscillation, corresponding to the duration of the timewhen the force is near its maximum. When taking this fac-tor into account, we estimate an “effective loading rate,” ora loading rate that would be comparable to loading rates ofnon-alternating force applications, of ≈
70 pN / s for our ex-periments.Our results indicated reproducible differences between thebiotin-coated surfaces and the control surfaces, in that the for-mer has a feature at a low rupture force. However, this forcevalue, estimated at 20(8) pN, is a factor of 2-3 lower than lit-erature values for rupture force of specific biotin-streptavidin,when compared to the rupture force at a similar loadingrate. The force range of the current setup will need tobe improved to study and resolve non-specific interactions. Atthe current stage, it is clear that our new detection system candistinguish different surface properties. In a separate experimental sequence, magnetic imagingdata were obtained. The excitation laser was positioned overthree bound particles, and the magnetic field image was ob-tained before, during, and after applying the pulses (describedpreviously) to the piezoelectric actuator (Fig. 7). The pulseswere paused at 40 V p − p in order to obtain the data for the “dur-ing" image, and the “after" image was obtained after rampingto 50 V p − p . In the magnetic images, blue and red regions cor-responding to positive and negative local magnetic fields withrespect to applied external magnetic field, are clearly visiblefor each of the three particles, indicating the direction of themagnetic dipole of each particle. The maximum measuredfield strength is about 0 . (cid:0) × − T (cid:1) , nine orders of mag-nitude larger than that typically detected when performingFIRMS using a vapor cell magnetometer (cid:0) × − T (cid:1) . The magnetic images have highest uncertainty along the edgesof the images where the laser intensity was weakest.
IV. DISCUSSION
There are several factors that may introduce variance inour observations. As previously discussed, the microparticlesused in these experiments have a broad size distribution (coef-ficient of variance of 0.2), which corresponds to a larger dis-tribution of applied forces. Such a distribution is expected toresult in an increase in the uncertainty of the measured ruptureforce.It is also possible that the mechanical motion of the cham-ber may not be accurately predicting the applied force. Com-pared to the rupture force, the variation in the conversion fac-tor (pN / V, Fig. 6) between applied voltage and applied forcewas found to be significantly smaller, at 10% - 30%. Thisleads us to believe that although the motion of the differentchambers is similar, the force experienced by the particles isnot.A possible cause of the variation in the apparent forcecomes from a violation of the assumption that the fluid inthe chamber is incompressible. Within 10 or 20 minutes ofthe sealing of the chamber, air bubbles can be observed alongthe circular edge of the chamber. These bubbles could allowfor the chamber to compress and expand with each oscilla-tion, altering the force experienced by the particles. An un-sealed open chamber would not be a good alternative due toevaporation during the course of the experiment. However,a microfluidic chamber with flow appears to be a promisingroute for future experiment, offering a stable configurationwith the benefit of reusability. We have also recently demon-strated success with a local manipulation of particles usingdielectrophoresis, which appears to be a promising methodfor future work.Magnetic imaging of the particles is an attractive alterna-tive to optical wide-field imaging in the presence of signifi-cant optical backgrounds or when optical access to the sampleis unavailable. The latter could be the case, for example, whenusing blood. Magnetic labeling has been demonstrated tobe a valuable tool for tracking biological entities. An illustra-tion of discriminating power of magnetic imaging can be seen White Images (0-255)
10 5 0 B e f o r e
10 5 0 D u r i ng P o s i t i on ( m )
0 5 10 15 10 5 0 A ft e r Magnetic Field (G)
0 5 10 15 Position ( m) -0.3-0.2-0.100.10.20.3
1 Uncertainty (G)
0 5 10 15 00.020.040.060.080.10.120.140.160.180.2
FIG. 7. Images of particles before (top), during (middle), and after (bottom) application of force. Wide-field optical images are displayed onthe left, where particles are visible as dark regions with Arago (Poisson) spots. Magnetic images are displayed in the middle with corresponding1 σ uncertainty displayed on the right. in the “during” and “after” images of Fig. 7: the particle at thebottom right is apparently non-magnetic, which is impossibleto tell from the optical image. Furthermore, magnetic imagesoffer information about the orientation of the particles, whichcan be used to discern whether a particle has remained at-tached, or has detached and quickly reattached to the surface.For example, when comparing the white images of the parti-cle in the middle of the “before” and “during” images shownin Fig. 7, the particle appears to move slightly to the right, butit is not known whether it is from sliding or rolling across thesurface. When we examine the corresponding magnetic im-ages and fit the orientation, we observe a spatial rotation witha total angle of 10(2) degrees.When compared against the original FIRMS technique, thismethod offers significantly (by nine orders of magnitude)boosted signal strength, allowing for the resolution of indi-vidual particles, in addition to in-situ force application. Thistechnique also allows for massively parallel experiments withvarious particles and/or surface functionalizations in a lab-on-a-chip environment. However, an important differenceis that the field-of-view of the current microscopy-based ap-proaches is much smaller, by two to three orders of magnitude,than those of FIRMS using an atomic magnetometer. Conse-quently, a much smaller number of molecular bonds are mea-sured, which may affect the statistics of the force distribution. V. CONCLUSION
We successfully realized a diamond-based FIRMS experi-ment capable of resolving individual particles with both op-tical and magnetic imaging, in addition to quantifying thestrength of the molecular interactions. Although we found thatwith our method of applying force, the apparent rupture forceof the particles varied significantly between chambers, thiscan potentially be addressed with an alternate method of forceapplications, such as dielectrophoresis, as previously demon-strated in our laboratory. This technique may prove useful indrug testing and diagnostics: we envision a system of multi-ple parallel microfluidic channels carrying nanoparticles withvarious functionalizations that are detected with high sensitiv-ity, and whose targeted molecular interaction is quantitativelymeasured.
VI. ACKNOWLEDGEMENTS
The authors acknowledge technical help and advice fromTao Wang, Andrey Jarmola, as well outstanding support withdiamond surface functionalization from Olga Shenderova, andMarco Torelli of Adámas Nanotechnologies. This work wassupported in part by the EU FET OPEN Flagship ProjectASTERIQS. SX acknowledges partial support by the NSF(1508845). V. T. Moy, E.-L. Florin, and H. E. Gaub, “Intermolecular forces and ener-gies between ligands and receptors,” Science , 257–259 (1994). M.-G. Kim, Y.-B. Shin, J.-M. Jung, H.-S. Ro, and B. H. Chung, “Enhancedsensitivity of surface plasmon resonance (SPR) immunoassays using aperoxidase-catalyzed precipitation reaction and its application to a proteinmicroarray,” Journal of Immunological Methods , 125–132 (2005). P. M. Yangyuoru, S. Dhakal, Z. Yu, D. Koirala, S. M. Mwongela, andH. Mao, “Single-molecule measurements of the binding between smallmolecules and DNA aptamers,” Analytical Chemistry , 5298–5303(2012), pMID: 22702719. M. Masetti, H.-n. Xie, Z. Krpeti´c, M. Recanatini, R. A. Alvarez-Puebla,and L. Guerrini, “Revealing DNA interactions with exogenous agents bysurface-enhanced raman scattering,” Journal of the American Chemical So-ciety , 469–476 (2015), pMID: 25496029. G. Sitters, D. Kamsma, G. Thalhammer, M. Ritsch-Marte, E. Peterman,and G. Wuite, “Acoustic force spectroscopy,” Nature Methods , 47–50(2015). L. De Silva, L. Yao, Y. Wang, and S. Xu, “Well-defined and sequence-specific noncovalent binding forces of DNA,” The Journal of PhysicalChemistry B , 7554–7558 (2013), pMID: 23734715. H. Jia, Y. Wang, and S. Xu, “Super-resolution force spectroscopy reveals ri-bosomal motion at sub-nucleotide steps,” Chem. Commun. , 5883–5886(2018). T. Ota, T. Sugiura, and S. Kawata, “Rupture force measurement of biotin-streptavidin bonds using optical trapping,” Applied Physics Letters ,043901 (2005). S. Wong, E. Joselevich, A. Woolley, C. Cheung, and C. Lieber, “Cova-lently functionalized nanotubes as nanometresized probes in chemistry andbiology,” Nature , 52–55 (1998). J. Wong, A. Chilkoti, and V. T. Moy, “Direct force measurements ofthe streptavidin–biotin interaction,” Biomolecular Engineering , 45–55(1999). S. B. van Dam, M. Walsh, M. J. Degen, E. Bersin, S. L. Mouradian, A. Gal-iullin, M. Ruf, M. IJspeert, T. H. Taminiau, R. Hanson, and D. R. Englund,“Optical coherence of diamond nitrogen-vacancy centers formed by ion im- plantation and annealing,” Phys. Rev. B , 161203 (2019). I. Fescenko, A. Laraoui, J. Smits, N. Mosavian, P. Kehayias, J. Seto,L. Bougas, A. Jarmola, and V. M. Acosta, “Diamond magnetic microscopyof malarial hemozoin nanocrystals,” Phys. Rev. Applied , 034029 (2019). Adámas Nanotechnologies. . K. Jensen, P. Kehayias, and D. Budker, “Magnetometry with nitrogen-vacancy centers in diamond,” in
High Sensitivity Magnetometers , edited byA. Grosz, M. J. Haji-Sheikh, and S. C. Mukhopadhyay (Springer Interna-tional Publishing, Cham, 2017) pp. 553–576. N. Bar-Gill, L. Pham, A. Jarmola, D. Budker, and R. Walsworth, “Solid-state electronic spin coherence time approaching one second,” Nature com-munications , 1743 (2013). D. Budker, “Diamond nanosensors: The sense of colour centres,” NaturePhysics , 453–454 (2011). E. Hecht,
Optics , Pearson education (Addison-Wesley, 2002) pp. 291–302. E. A. Evans and K. Ritchie, “Dynamic strength of molecular adhesionbonds.” Biophysical journal , 1541–55 (1997). R. Merkel, P. Nassoy, A. Leung, K. Ritchie, and E. Evans, “Energylandscapes of receptor-ligand bonds explored with dynamic force spec-troscopy,” Nature , 50–53 (1999). L. Yao and S. Xu, “Force-induced remnant magnetization spectroscopy forspecific magnetic imaging of molecules,” Angew. Chem. Int. Ed. , 4407–4409 (2011). M. Kayci, S. Lourette, T. Wang, L. Bougas, and D. Budker, “Quantitativemeasurements of non-covalent interactions with diamond based magneticimaging,” Applied Physics Letters , 053103 (2018). C. Bustamante, Z. Bryant, and S. B Smith, “Ten years of tension: Single-molecule DNA mechanics,” Nature , 423–7 (2003). Y.-T. Chen, A. Kolhatkar, O. Zenasni, S. Xu, and T. Randall Lee, “Biosens-ing using magnetic particle detection techniques,” Sensors , 2300 (2017). H. Liang, X.-B. Zhang, Y. Lv, L. Gong, R. Wang, X. Zhu, R. Yang, andW. Tan, “Functional DNA-containing nanomaterials: Cellular applicationsin biosensing, imaging, and targeted therapy,” Acc. Chem. Res.47