Portable NMR Detector by fcFthw

VIEWS: 76 PAGES: 24

									Portable NMR Detector


      May 29, 2005
      Ryan Hubbard
        David Ng
         Relevance of Work
• Current NMR/MRI
  devices are bulky and
  expensive
• Impossible to use in
  the field
• Most students never
  get hands on
  experience
            Project Goals
• Create a small, relatively inexpensive,
  detector compatible with current NMR/MRI
  instrumentation
• Device must be small and light enough for
  one person to carry
• Device should have sufficiently small line
  width to resolve different chemicals
  (10ppm)
The Science of Spin
          • Protons act as tiny
            magnets
          • When an external field
            (B0) is applied,
            protons align with or
            against the field
          • A slight majority align
            with the field, so there
            is a net magnetization
            in the direction of B0
          The Science of Spin
• By applying an orthogonal,
  oscillatory field with an
  inductor, we can tip the
  magnetization into the plane
  of the inductor.
• This vector then precesses
  like a gyroscope, creating a
  changing net magnetic field.
• By Faraday’s Law, this           ω = γB0
  creates a voltage signal in
  our inductor.                  γ = 42.58 MHz/T
                                    for protons
           Design: Instrumentation
• Originally, we planned to
  use a Techmag Apollo
  controlled with NTNMR
  software under WIN2k.
  Under testing, this proved
  very unreliable.
                               • Instead, we moved to a
                                 Varian system controlled
                                 with its own software
                                 running on a Sun
                                 platform.
                  Design: Switch
• We use the same coil to create
  our orthogonal field as we do to
  receive the signal, so a
  transmit/receive switch is
  needed
• Initially we used crossed diodes
  and a λ/4 cable, but the signal
  bled through our diodes,
  causing artifacts
• The Varian system has an
  integrated switch, which only
  requires the addition of a proper
  λ/4 cable
Design: RF Coil
        • Induces precession of
          protons, and detects
          generated signal
        • LC Resonator, with
          variable capacitors for
          impedance matching
          and tuning
        • Tuned to frequency of
          precession. For the .64
          T magnet ~ 27 MHz
            Design: Magnets
“Little Boy”
• ~.6 T B0
• Four layers of 8 NdFeB
  magnets in a circular
  Halbach array

                           “Fat Man”
                           • ~1.1 T B0
                           • 36 magnets in 18 rods
                             forming a hexagonal
                             Halbach array
           Fabrication: RF Coil
• The inductor was hand wound,
  then soldered into place on a
  PCB along with non-magnetic
  capacitors and the co-axial
  cable
• The coil was then manually
  tuned to a frequency slightly
  below the frequency of
  precession using a spectrum
  analyzer
• A free-space coil was
  constructed for “Fat Man”, to
  eliminate any possible noise
  caused by the large copper PCB
              Fabrication: “Little Boy”
• Four magnet “sandwiches,” 8
  magnets per layer
• All 32 magnets characterized and
  specifically placed to optimize
  homogeneity
• 4 magnets were epoxied onto each
  plate, and the two plates were then
  epoxied together using bolts to keep
  them centered
• Finally, the four layers were stacked
  together to create the whole magnet.
• Teflon runners epoxied to the brass
  bore plating were used to mount the
  coil inside the magnet.
Fabrication: “Fat Man”
            • Each magnet characterized as
              before
            • Two matched magnets aligned
              and epoxied into each tube
            • Each tube specifically placed
              to optimize efficiency
            • Tubes aligned using notched
              caps on each end, and held in
              place using set screws
            • Unfortunately, magnets are
              stronger than the aluminum
              keys used to align them,
              resulting in breakage
                   Testing: Factors
•   Pulse Width Adjustment: Is the
    signal really an FID?
•   Various Sample Materials: Can the
    detector differentiate substances?
•   Various Sample Sizes: How do the
    dimension of the sample affect
    amplitude and line width?
•   Power Adjustment: How do
    amplitude and line width change?
•   Coil Position: How homogenous is
    the detector?
•   Tolerance: How do changes in coil
    tuning affect output?
                   PW Adjustment
 • Varying the pulse width adjusts the degree to which the
   net magnetization is projected onto the XY plane




• This is sinusoidal, with the magnetization tipping further onto the
  XY plane, and then rotating further past it.
                 Various Materials



Ultrasound Gel    Epoxy Pt. A         Epoxy Pt. B          Vaseline




 Empty Plastic Chamber          Air                 Teflon Stage
           Various Materials
• Line width is too broad for proton rich samples, so
  different materials cannot be differentiate
• The 50% bandwidth of 5 kHz indicates a ΔB0 of
  11.74 G
• However, signal amplitude is an order of
  magnitude lower in samples without protons
• Essentially what we have is a proton counter
• Note that there is a constant frequency artifact
  near 27.07 MHz due to noise
Power Comparison
                 • Power adjustments do
  Higher Power     not seem to have an
  Full Sample      appreciable effect on
                   signal amplitude or
                   bandwidth
                 • Signal amplitude is
                   directly proportional to
                   the net magnetization
  Higher Power     due to B0, so it should
  Small Sample     be independent of the
                   RF pulse
                    Sample Sizes
                                                         Effect of Sample size on Bandwidth and Signal Amplitude

• Amplitude                                   1.20E+06                                                         3

  approximately
  proportional to signal                      1.00E+06                                                         2.5


  size
                                              8.00E+05                                                         2
• Bandwidth decreases      Signal Amplitude


  for smaller samples,




                                                                                                                     Bandwidth(kHz)
                                                                                                                                      Amplitude
                                              6.00E+05                                                         1.5
                                                                                                                                      Bandwidth
  but also geometry
  dependent                                   4.00E+05                                                         1



• Lower limit on
                                              2.00E+05                                                         0.5
  bandwidth appears to
  be 1.57 kHz, ~ 58ppm                        0.00E+00                                                         0
                                                         Tiny       Small     Medium       Half       Full
                                                                            Sample Size
              Homogeneity
• Signal is strongest in   Disp.    BW                 Center
                           (mm)    (KHz)     Peak      (MHz)
  the center of the
                            0      5.8824   9.30E+05    27.114
  magnet (Highest B0)
• It is slightly more
                            1      5.4902   9.12E+05    27.116

                            2      5.4902   8.38E+05   27.11697
  homogenous at 3mm
  away from the center,     3      3.5295   5.05E+05   27.1174

  but much worse            4       13      1.50E+05   27.1156

  outside that range        5       20      7.50E+04    27.11

                            6       23      5.90E+04    27.095
                        Tolerance
 • Decreasing peak attenuation by
   adjusting variable capacitors
   had no significant effect on
   results, once again because
   signal strength is mostly
   dependent B0
                                               Peak attenuation -10dB
Out of Tune      Amplitude          • Detuning the coil by adjusting
   0 kHz           1.07E+06           variable capacitors to shift
   25 kHz          8.64E+05           frequency reduced signal
   50 kHz          8.29E+05           strength, because fewer protons
   75 kHz          8.11E+05           are precessing at this frequency.
  100 kHz          7.80E+05
       Future Improvements
• Fat Man – Close to Clinical MRI Strength
• Shimming – Increase Homogeneity
• Compact Instrumentation – True Portability
• Free Space Coil – More Stable System
• Gradient Coils – Imaging Applications
• Spin-Echo Sequences – Negates some in-
  homogeneities in imaging
• Shielding Against Noise
                 Thanks
• Dr. Luisa Ciobanu, Dr. Boris Odintsou, Dr.
  Andrew Webb, Mary Jane Ham, Mikhail
  and everyone at BIC for use of their
  facilities and experience
• Scott Mc Donald and Scott Sprague for all
  our machine work
• Frank Dale and the ECE Shop for parts
• Shenghui Zhang for general assistance
                 References
• J.P. Hornak “Spin Physics” The Basics of NMR (J.P.
  Hornak, 1997)
• G. Moresi & R. Magin “Miniature Permanent Magnet for
  Table-top NMR” Concepts in Magnetic Resonance Part B,
  Vol. 19B(1) pg. 35-43 (Wiley Periodicals, 2003)
• H. Raich, P. Blümler “Design and Construction of a
  Diploar Halbach Array with a Homogeneous Field from
  Identical Bar Magnets: NMR Mandhalas” Concepts in
  Magnetic Resonace Part B, Vol. 23B(1) pg. 16-25 (Wiley
  Periodicals, 2004)
Questions?

								
To top