Integrated gradiometer

#The present invention relates to an integrated sensor and method for measuring magnetic field gradients. More particularly, the present invention is directed to sensing and measuring magnetic field gradients that spatially vary on a length scale of 10 microns or less, such as are generated by microscopic magnetic particles that may be bound to a target species for identification.

#There is an increasing demand for the ability to detect magnetic fields that vary on a small spatial length scale of d μm, where d may vary from the order of 0.01 microns (10 nm) to the order of 10 microns. As one particular example, fields that vary on such a scale are associated with the magnetic moment of magnetic particles (MPs) of diameter d. One method for detecting the presence of such a particle is to sense the accompanying, local magnetic field. These particles, in turn, may be used for a variety of applications, such as “tagging” biological or chemical targets. Thus, sensing the presence and/or location of magnetic “tags” permits the identification of the associated target. The background art essentially involves two categories, the use of individual sensors to detect local magnetic fields, and common gradiometric techniques used for measuring field and field gradients that vary on a macroscopic scale.

#Regarding the first category, the prior art describes the use of individual field sensors for the specific application of detecting magnetic particles. Criteria for suitably designed sensors are known to those skilled in the art. It is generally desirable to fabricate a magnetic field sensor that can be located in close proximity to the particle, within a length scale of order d, and furthermore the size scale of the sensor itself must be of order d. U.S. Pat. No. 5,981,297, Baselt, describes a scheme using magnetoresistive sensors for the specific application of detecting magnetic particles that are used to tag biological agents. U.S. patent application Ser. No. 09/497,754, filed Feb. 4, 2000, by G. A. Prinz and M. M. Miller, describes a specific type of sensor configuration and teaches critical details for realizing such a sensor. In both these disclosures, Giant Magnetoresistance (GMR) sensors are a preferred embodiment. The advantage of GMR is its relatively large magnetoresistive effect and the associated high sensitivity that, in general, surpasses that of Hall effect devices. However, the sensitivity of a GMR element is compromised when the device is fabricated on a small scale because of magnetostatics. For example, a sensor that is many microns in size may be capable of detecting magnetic fields that are uniform over the sensor in the nT range. However, the fringing field from a MP that is locally highly non-uniform becomes problematic for several reasons. First, if the active area of the sensor is much larger than the MP then the magnetoresistance induced from the fringing field is largely shunted out by the unperturbed portion of the sensor. Furthermore, if the MP is extremely small the highly localized fringing field will be unable to affect a comparably small portion of the sensor. This is because the magnetic moments in the sensor cannot be arbitrarily rotated at a greater spatial variation than that allowed by the exchange interaction. Second, when the GMR sensor has dimensions on the order of d the micromagnetic configuration of the sensor becomes the dominant concern as d becomes very small. Any conceivable GMR sensor element will have to overcome magnetostatic demagnetization barriers in either the “quiescent” state or when the magnetic moments are either reversing or rotating in response to the MP. For many reasons, fabrication of GMR devices become extremely difficult for length scales d less than 1 micron.

#FIG. 1 is a schematic, cross-sectional view of a magnetic particle and the dipolar, fringe magnetic fields associated with the particle.

#FIG. 2 is a top view of a preferred embodiment of an integrated gradiometer using a vertical stack arrangement of Hall cross sensors.

#FIG. 3 is a cross-sectional view of the embodiment of FIG. 2.

#As used herein, the term “local magnetic field source” refers to a microscopic magnetized object that generates a local magnetic field with an associated field gradient. An example is a microscopic magnetic particle, with diameter of order 10 nm up to 5 microns. Where reference is made herein to first and second sensors that are spaced apart at a distance effective for measuring a magnetic field gradient having a spatial scale of d microns, the meaning is as follows. The locally strong magnetic fields associated with microscopic magnetic particles, or generally any microscopic magnetic structure, can be modeled as coming from a superposition of sources characterized as monopole or dipole. Typically, components from dipolar sources dominate the characteristics of the local field. The spatial scale d that defines the dipole is a characteristic length scale for the spatial variation of the local magnetic field. It has been noted that the lateral dimensions of the first and second sensors should be d or smaller, to avoid averaging the field over an area so large that spatial resolution is lost. The separation between the first and second sensors should also be of order d. If the separation is larger than d, one sensor acts essentially as a null detector. This size scale can be used, for example, when an external field is applied in order to increase the moment of a paramagnetic source. The detection of a particle is permitted, but there is no spatial resolution of the field gradient. A separation of less than d, for example d/4 or d/2, is optimal for a quantitative measurement of the field gradient. As an example, measuring the field gradient of a magnetic particle with diameter of 2 microns would employ two sensors spatially separated by a distance of ½ to 1 micron.

#FIG. 1 shows a magnetic particle MP 10 with diameter d. In some cases, d represents the full diameter of the particle and, in other embodiments, d includes a coating or membrane that functionalizes the MP. Its magnetic moment (m) 12 may be a spontaneous moment or it may develop in proportion to an externally applied magnetic field. If the moment is spontaneous the orientation of the moment may be random, but may be manipulated by local or external fields. If the moment is induced using an externally applied magnetic field, the moment will typically align with this field with the strength of the moment being proportional to the applied field. Below we will discuss these issues in greater detail. It's reasonable to approximate the moment as a dipole, and the fringe magnetic fields associated with m are also sketched in FIG. 1. The important point is that the dipolar fringe fields 14 have a strong spatial variation on the length scale d.

#Very near the surface of MP 10, the fringe field 14 magnitude has approximate value B(r=d/2)˜4πM (M is the volume magnetization, M=|m|/V) which may be the order of 0.01 to 1 tesla. Because of its dipolar nature, B˜(1/r3), the fringe field 14 decays by orders of magnitude within a distance of several d. It follows that (i) the field gradient δB/δr is large, and (ii) the particle's fringe field may become undetectable at distances greater than a few d. Thus, it is generally desirable to fabricate a field sensor that can be located in close proximity to the particle, within a length scale of order d, and the size scale of the sensor itself must be of order d.