Bio-NMR

June 30-July 1, 2008

Overall goals

Be able to set up a sample for quality data collectionBe able to use vendor-supplied parameter sets and pulse programsBe able to run the instrument safelyImprove understanding of how the instrument works

Probe designSpecial purpose probes emphasize a few aspects of performance at the expense of othersGeneral purpose probes are compromisesEverything gets more difficult at high fieldEverything gets more difficult in a cryoprobe

Probe capabilities/performance

Field homogeneity/lineshapeRadio frequency coilsGradient coilsTemperature control

Probe design issues

Magnetic susceptibility considerationsAcoustic ringingBackground signalsMechanical stability and robustnessWeightDisassembly for cleaning and repair

Practical RF coils

Helmholtz shapeLimitations on the uniformity of excitation inside the coil, restriction of field outside the coil, degree of inversionCoil’s magnetic susceptibility must be masked

Coil layout

Generally limited to two coils, often four RF channels“Triple” inverse probes for bio H on the inside C and N (or C and P) outside D either inside or outside

Electrical performanceTuning is adjustment of circuit resonant frequency to NMR frequencyMatching is impedance matching of the circuit to the pulse amp output and preamplifier inputDielectric mostly changes tuning, ionic strength mostly changes matchingIons in solution degrade the electrical performance

Gradient coilsGradients used for coherence selection, artifact control, shimming, diffusionReproducibility of gradient pulses is of highest importanceLinearity is not perfectZ is most important, x and y usefulFor shimming, x and y room temp shim coils can be ramped

Temperature controlThermocouple cannot be placed inside RF coilsTall, thin tubes maximize temperature gradients; very complex behaviorWatch out for cancelling a temperature gradient with a homogeneity gradientConvection cells in the tube are really badTemperature limits fixed by hardware

Factors influencing S/N

Number of nuclei presentEfficiency of the coilCoil’s quality factor (Q)Sample geometryIonic strength of the sample

Power handling

Risk from high power is excessive voltageRisk from long pulses/decoupling is coil/sample heatingCooling is critical during decoupling

Ways of describing an RF pulse

Voltage VPower, W (P=V2/R) Decibels, dBDuration, us/ms/sPhase/phase cyclePurposeTip angle, degrees (proportional to time x voltage x profile)Field strength, Hz (gammaH1/2pi)

Decibel scale for voltage and power

Log scaleCan be either relative or absolutedB=20 log(Vin/Vout) (proportional to pw)

dB=10 log(Pin/Pout)

Change 90 degree pulse width by 10x =20 dBChange 90 degree pulse width by 2x

= 6dB

Effects of pulses at various power levels

Excitation (full power, usually 90)Inversion (full power, 180)Refocusing (full power, 180)13C, 15N decoupling (down 10-12 dB)Spin lock (down 15-20 dB)1H decoupling (down 18-20 dB)Water flipback (down 30-35 dB)Water presaturation (down 55-75 dB)

Varian and Bruker dB scales

Varian runs - (min) to 63 (max) in coarse steps of 1 dB, plus a fine power adjustment in arbitrary DAC units Typical rectangular pulses 55-60 Typical presaturation 6

Bruker runs 120 (min) to –6 (max) settable to at least 0.01 dB Typical rectangular pulses 0- -6 Typical presaturation 55

Pulse shapingSoftware allows creating pulses of arbitrary shapePulses can be optimized for a particular property (at the expense of others)Pulses can be selective for certain frequencies (a.k.a. structural type), or made very broadbandCalibrating pulses at one power/shape can be used to predict calibration for other powers and shapes

The most important thing to remember about what shaped pulses do!

The shape of a pulse in the time domain and its profile in the frequency domain are related by a Fourier transform Exponential: Lorentzian Rectangular: Sinc Gaussian: Gaussian

Profiles of shaped 90 pulses

Profiles of shaped 180 pulses

Predefined power levels/shapes for triple resonance

Nitrogen and deuterium: 90 degree decoupling

Proton: 90 degree Decoupling Tocsy spinlock Roesy spinlock Flipback Presat

Pulse shapes for carbon

90 degree rectangleDecouplingAdiabatic inversionAdiabatic refocusingAlpha or carbonyl selective 90 (also “time reversed”)Even more selective 90

Alpha and carbonyl 90 (also “time reversed”)Alpha/carbonyl 180Alpha or carbonyl decouplingAlpha and carbonyl decouplingAdiabatic decoupling (two power levels)Duplicate lists for carbon as decoupler and direct observe

Shaped gradient pulses

Original Bruker scheme: sine pulses, very softOriginal Varian scheme: rectangular pulses, very hardNew schemes: Bruker=smoothed square; Varian=WURST; very similar in practice

RF channel configurations

Our 400’s and new 500: 2Existing 500: 3600 and 800:4 For proton observe experiments,

channel 1 is proton Channel 2 is carbon Channel 3 is nitrogen Channel 4 is deuterium

Spectrometer evolution

20th century spectrometers: frequency synthesizer sources, lots of components needed to adjust the gating, amplitude, and phase of pulses21st century spectrometers: direct digital synthesis, far fewer components, each running their own software (with their own bugs)

RF amplifier considerations

All modern spectrometers use linear amplifiersGain typically 60 dBMaximum power 100-500 wattsLong linear range, compression near top end of power outputSoftware incorporates some type of correction for nonlinearities

Safe power handlingProbes come with a specification sheetNewer Bruker probes have a memory chip containing information on limitsSoftware power controls Current ones do not cover every scenario

Cryoprobes handle power much more efficiently, but can accept less; overall, pulse widths are longer than in conventional probes

Streamlining setup of many experiments

A pulse sequence using a consistent nomenclature for pulses and delaysA parameter set that contains all the information specific to the experiment, e.g. sweep widths, delay lengths, gradient powersCalibration of local probes and amplifiers in a probe fileMerge the three and go

The Varian way—BioPackPulse and delay names managed by George Gray at VarianParameters accessed through drop down menusA probe file called HCN is managed out of the gHNCO parameter setAn extensive set of auto-calibration and auto-acquisition tools

The Bruker way—rpar, getprosol

Rpar = read global experiment parameters (and pulse sequence)Getprosol = read probe and solvent specific parametersProbe file managed through edprosolEvery pulse sequence references a relations file

Pulse programming issuesVarian pulse programs written in a C-like language (possibly with a visual editor) Extensive use of flags to control optional

features Tends to minimize the number of programs

Bruker pulse programs written in a machine-like language Most features are hard coded in the

sequence Tends to maximize the number of programs Bruker-supplied programs follow a specific

naming convention

Vendor-specific differences

Varian favors fewer gradient pulses, Bruker favors moreVarian favors States-TPPI scheme for phase sensitive detection, Bruker favors echo-antiechoBruker programs mostly written by one person, Wolfgang Bermel; Varian programs come from many sources; several customer labs contribute

Sample specific parameters

The proton 90 degree pulse in an inverse probe is very sensitive to the composition of the sample because proton is on the inner coilThe carbon and nitrogen pulses are much less sensitive, because they share the outer coil

Accessing help and documentation

Varian Global manuals on Help menu Every BioPack sequence has a text

manual

Bruker Global manuals on Help menu NMR Guide and Encyclopedia Terse but helpful hints in the comments

section of the pulse program

Power limits

High power is limited such that proton pw90 >6, carbon >15, nitrogen >35 in Varian probeProton >8, carbon >16, nitrogen >40 in Bruker probeDecoupling power limits are reduced as at/aq is increased

Duty cycle considerations

Maximum duty cycle is 8% with conventional garp decoupling at ~10 wattsTo a first approximation, duty cycle is at/(d1 + at)“Canned” parameters will be safe with d1=1 and at~0.08 secModified parameters may not be safeRefer to the document “Cp800.pdf”

Lower power decouplingVarian prefers adiabatic decoupling on nitrogen and carbon channels at about 60% of the average power of garp decouplingBruker parameterizes many experiments for adiabatic carbon decoupling, some fast (“BEST”) nitrogen experiments for garp4 decoupling at 25% of normal power (pulse width apx. doubled, power down 6 dB)Garp4 works for carbon too and we can use simultaneous garp4 decoupling for both nitrogen and carbon

Temperature limits and sample positioning

Cold probes have a limited temperature range (Bruker 10-60, Varian 0-50)Actual temperature is a few degrees colder than thermocouple readingTemperature gradients a bigger problem than in warm probesBruker probes have a solid, and FRAGILE, bottom

Radiation damping and water suppression

What works in warm probes at low field may not work, or make things worse, in cold probes at high fieldRecalibrate flipback pulsesZero out any trim pulses in hsqc type experiments