9 May 2008 | J. M. Taylor1*, P. Cappellaro2,3*, L. Childress2,4, L. Jiang2, D.Budker5, P. R. Hemmer6, A.Yacoby2, R. Walsworth2,3, M. D. Lukin2,3
A high-sensitivity diamond magnetometer with nanoscale resolution is presented, utilizing Nitrogen-Vacancy (NV) centers in diamond for detecting weak magnetic fields. The NV centers, which can be individually addressed, optically polarized, and detected, offer excellent coherence properties even at room temperature. The magnetometer can detect the precession of single nuclear spins and provide optical magnetic field imaging with spatial resolution ranging from micrometers to millimeters and sensitivity approaching a few femtotesla per square root of hertz.
The detection of weak magnetic fields with high spatial resolution is important in various fields, including fundamental physics, material science, data storage, and biomedical science. Various magnetic sensors have been developed using techniques such as superconducting quantum interference devices (SQUIDs), the Hall effect in semiconductors, atomic vapor and BEC-based magnetometry, and magnetic resonance force microscopy. This paper presents a novel approach using solid-state electron spin quantum bits, specifically NV centers in diamond, for high spatial resolution magnetic field detection.
The operating principles of the magnetometer are closely related to those of magnetometers based on spin precession in atomic vapors. Detecting the relative energy shift induced by a magnetic field between two Zeeman sublevels allows for precise determination of an applied DC or AC magnetic field. The sensitivity is determined by the spin coherence time and spin projection noise. Although solid-state electronic spins have shorter coherence times than gaseous atoms, quantum control techniques can decouple them from the local environment and from each other, leading to improved sensitivity.
The magnetometer uses a Ramsey-type sequence to detect the Zeeman shift. A π/2 pulse creates a superposition of two Zeeman levels, which acquire a relative phase from the external field during the free evolution interval. Another π/2 pulse transforms the relative phase into a population difference, which is measured optically and used to infer the Zeeman shift. For small phases, the magnetometer signal is linear in the magnetic field. Increasing the interrogation time improves the sensitivity until environmental perturbations lead to decay of the free-precession signal.
In solid-state spin systems, coherence is limited by interactions with nearby lattice nuclei and paramagnetic impurities, resulting in an ensemble dephasing time. Additionally, there is a finite number of fluorescence photons collected and detected, leading to additional photon shot noise and a finite contrast to the Ramsey fringes. These effects are described by a single parameter C ≤ 1, which approaches unity for ideal, single-shot readout.
The sensitivity of the magnetometer based on a single electronic spin is given by η_DC ≈ ħ/(gμ_B C√T_2*). For current experiments with detection efficiency ~10^-3, C ≈ 0.05, and T_2* ~ 1 μs, the optimal sensitivity is ~1 μT/Hz^1/2. Improving the collection efficiency to ~5%A high-sensitivity diamond magnetometer with nanoscale resolution is presented, utilizing Nitrogen-Vacancy (NV) centers in diamond for detecting weak magnetic fields. The NV centers, which can be individually addressed, optically polarized, and detected, offer excellent coherence properties even at room temperature. The magnetometer can detect the precession of single nuclear spins and provide optical magnetic field imaging with spatial resolution ranging from micrometers to millimeters and sensitivity approaching a few femtotesla per square root of hertz.
The detection of weak magnetic fields with high spatial resolution is important in various fields, including fundamental physics, material science, data storage, and biomedical science. Various magnetic sensors have been developed using techniques such as superconducting quantum interference devices (SQUIDs), the Hall effect in semiconductors, atomic vapor and BEC-based magnetometry, and magnetic resonance force microscopy. This paper presents a novel approach using solid-state electron spin quantum bits, specifically NV centers in diamond, for high spatial resolution magnetic field detection.
The operating principles of the magnetometer are closely related to those of magnetometers based on spin precession in atomic vapors. Detecting the relative energy shift induced by a magnetic field between two Zeeman sublevels allows for precise determination of an applied DC or AC magnetic field. The sensitivity is determined by the spin coherence time and spin projection noise. Although solid-state electronic spins have shorter coherence times than gaseous atoms, quantum control techniques can decouple them from the local environment and from each other, leading to improved sensitivity.
The magnetometer uses a Ramsey-type sequence to detect the Zeeman shift. A π/2 pulse creates a superposition of two Zeeman levels, which acquire a relative phase from the external field during the free evolution interval. Another π/2 pulse transforms the relative phase into a population difference, which is measured optically and used to infer the Zeeman shift. For small phases, the magnetometer signal is linear in the magnetic field. Increasing the interrogation time improves the sensitivity until environmental perturbations lead to decay of the free-precession signal.
In solid-state spin systems, coherence is limited by interactions with nearby lattice nuclei and paramagnetic impurities, resulting in an ensemble dephasing time. Additionally, there is a finite number of fluorescence photons collected and detected, leading to additional photon shot noise and a finite contrast to the Ramsey fringes. These effects are described by a single parameter C ≤ 1, which approaches unity for ideal, single-shot readout.
The sensitivity of the magnetometer based on a single electronic spin is given by η_DC ≈ ħ/(gμ_B C√T_2*). For current experiments with detection efficiency ~10^-3, C ≈ 0.05, and T_2* ~ 1 μs, the optimal sensitivity is ~1 μT/Hz^1/2. Improving the collection efficiency to ~5%