Omenetto Cume Flashcards
What are thw two principial goals of analysis
Sensitivity and selectivity
What has development enabled new SAD methods and at the time are they practical?
High power wavelength tunable lasers - and they are not - not used for routine use but the paper wants to extrapolate their applciation
What are the major issues with SAD
Low efficiency of breaking something down into atoms and then distinguishing that signal from noise or interfering signal (so again sensitivity and selectivity)
This paper makes distinguishment between theoretical and actual LOD - why?
the current most sensitiite for SAD is ICP or electrically generated discharges - but these are at theoretical limits whereas LEAF laser excited atomic fluorescence is not and could theoretically do much better SAD
What are the two major classes of atom probing methods looked at in terms of SAD
Methods based on detection of charged species and methods based on fluorescence absorption and emmission
What is the key step in detection of charged based species and what are some examples given:
IONIZATION! some examples are Resonance ioniztn and laser enhanced ionization
What is the key step in detection of charged based species and what are some examples given:
IONIZATION! some examples are Resonance ioniztn and laser enhanced ionization
What does the figure on RIS vs LEI(OGE) show (and what is OGE)
THe figure shows the two ionization methods RIS vs LEI and more importantly how energy is gained and lost
RIS is two vertical - pointing up straight lines indicating radiational transition between energy states (absorption or pumping - if downward arrow indicates loss of photon through emission or fluorescence)
LEI (LASER!)has a vertical upward arrow but then a wavy arrow
THE WAVY indicates a thermal mode of gaining or losing energy - so upwards is thermal activation downwards is loss of energy such as collisional cooling
What is the optogalvanic effect
The Optogalvanic effect is the change in the conductivity of a gas discharge induced by a light source (typically a laser). This effect has found many applications in atomic spectroscopy and laser stabilization.[1]
Describe RIS and what is it and what are some key characteristics/issues
resonance ionization - basically hit with rapid succession of photons until ionized
since needs to be rapid - very improbably unless using a laser (in fact more than one laser)
Issue: NEEDS TO BE super clean sample chamber free from extraneous ions so at the time not feasible for real world dirty samples
What methods are capable of SAD?
RIS, resonance and non resonance fluorescence detected under energy level saturation conditions (LEAF -laser excited atomic fluorescence)
Atomic fluorescence generated by a laser in a carbon furnace atomizer
microwave induced plasma coupled with microarc atomizer and atomic absorption with graphite furnace atomizer
What is LEI and describe it strength sweaknesses
Laser enhanced ionization - a combination of thermal and irradiation -
atoms in a flame are always partially ionized so can increases percentage ionzied through irradiation
so keep atoms in a hotcell and shoot laser at it - detection is done through monitoring resistance or conductivity
AMENABLE to use in flames or plasma etc -
HIGH SENSITIVITY and high selectivity
NOT BELIEVED TO BE ABLE TO DO SAD
How does the author compare ionization to atomic absorption emission fluorescence techniques
AAES are Much more common - widely adapted in the world at the time - more typical lab analysis
What are the 3 fluorescence methods and describe them (energy level methods not analytical)
So resonance, direct line and stepwise
Resonance is straight up and down a solid line -
Direct line is straight up then a straight line down partially - emission - then collisional cooling the rest of the way
and step wise is straight up then collision cooling partially- then emission the rest of the way down
How are the 3 fluorescnece methods split up
REsonance vs NON resonance - resonance is just absorption from ground to excited and then emission (OF A RESONANT PHOTON OF SAME ENERGY ABSORBED)
Non resonance - emission of a photon OF AN ENERGY AND WAVELENGTH DIFFERENT FROM THAT ABSORBED (stepwise and direct line>
How can once increase fluorescence sensivity? and how do incident light levels play a role
Use a laser - at low incident light levels - the fluorescence is directly proportional to intensity in laser power
AT HIGH VALUES - can induce saturation
What is SATURATION in this regard (atomic energy level saturation)
basically the upper and lower enerhgy states are equalized at higher laser spectral irradiance (UNITS!)
the basis is when laser radiation is incident- a fraction is raised to upper state and this fraction increases as laser irradiance increases
The energized population can undergo another way to grounds state called SIMULATED EMISSION - identical to absorption but in a toward transition (eg resonant photon encounters an atom in an excited state) - if enough happens - upward and downward transition is equally probable
A
Why is LEAF a contender for SAD detection
Beacuse in energy level saturation - radiation based transition up and down GREALTY outweigh any mechanism making the excited state population determined entirely by the atoms intrinsic properties as opposed to the environment or anything else - this makes it a lot more controlled and stable /predictable
Also sources of error/variation such as quenching or laser power variation are minisucle compared to amount of resonance fluorescence
Whats the diagram for absorption and esmission events look liek
absorption straigh tup, emssiion is up wavy and down straight down
What are some absorption emission tehcniques and why shouldn’t they be as sensitive
ICP, Microwave induced plasma, and dc plasma - they cannot produce energy level saturation in the atomic emission process - thermodynamically impossible to inject enough thermal energy to sturate atomi energy levels - the temp required would be infinite
Whats the problem with atomic absorption methods
shold be less sensitivie they require measruement of small differences between two relatively large light levels P0 and P - drift or noise from the light source then precludes detection of anything approaching single atom level
So P0 and P are the power of radiation coming from the lamp or source and P is it post sample
These are too large light levels the natural drift and noise of these large levels absolutely precludes the detection of anything single atom level
What is the issue with saturated non resonance LEAF for SAD on real samples
IN ability to completely atomize the sample
also distinguishing signal from various sources of background from more conventional atomization technique eg incandescent carbon furnace - or flickering of laser
How does the author believe we can make LEAD for SAD more applicable
observing the atom over a longer period of time (should produce 10^8 photons per second)
What are the two atomization schemes
Both involves making a solution then 1) is sprayed into a flame or plasma (droplets must be evaporated deolated to free particles which then must vaprozie to make ions) or 2) dried as a solute in a furnace or heated cuvette and then thermally varpozied (ELECTROTHERMAL ATOMIZER)
so itemized - solid samples -> solution -> either spraying and desolvation or just drying -> vaporization -> atomization (sample vapor is atomized)
Author proposes 2 ways to increases effiency of atomization what are they
1) understand every process and optimzie
2) Brute force - high enough temp so each sample 100% atomized
3) find a new high efficiency atomization which gentle action that can thoroughly atomize
In which context does the paper talk about understanding atom formation
_ HE IS ONLY FOCUSING ON THE INVESTIGATIONS IN HIS LABORATORY WHICH IS ATOM FORMATION IN ANALYTICAL FLAMES AND PLASMA
Explain the authors conclusions on optimize atomization parameters (including figure 6)
In understanding atomization with plasma and flame - an issue is that aerosols are uneven so they created a set up in which one droplet is sent at a time and took a single flash photograph showing the single stream being sent to an acetylene torch - (rate of 1k/second) - due to stability can treat this as a time lapse even though different droplets Can see it moves to flame, evaporate as they move upward and solute particle vaporizes leaving luminescent plume -Ultimate conclusion though is even though can visualize each step - don’t know enough about it to assess/manipulate - this is for the future - long term goal
How does ICP work (review figure 7)
inductively coupled plasma is created by a coil around a quartz tube containing plasma gas (usually argon). t
1)through inductive coupling - the coil generates a field that
2)accelerates ions in the gas which collide with argon atoms and ionzie them which causes electrons to be released.
3)The electrons are accelerated to produce further ionization
4) final result is plasma with max temp of 10k C into which we can spray our sample
What is the problem with brute force atomization
The calculated energy to atomize a molecule is not very high but the effective temperature needed for it is very high - considerable - they point to probably the most successful as using thermally hot sources for atom formation and excitation - particularly ICP
Figure of Merit for ICP
Strength - atomization relatively complete due to high temp
Leads to high sensitivity and low interferences (small)
High temp at which emission generatedalso helps sensitivity
DOwnsides:
1 Spectral background very high due to hot plasma gases
2 Due to high heat a variety of elements are ionized including alkali and alkaline earth metals which have strong emission spectra which are scattered within in the monochromator or spectrometer used to separate elemental emission lines - and this scattering causes erroneous signal (this requires a high resolution detection system to distinguish this emission)
3 The source has flickering noise which causes variations in the atomic elemental signal and background (difficult to detect single atoms)
4 Large dilution with argon gas necessary (15-20 mL/min) - 1 ug/mL of sample reduced to 10^10 atoms/cm^3 in plasma - of course an issue for SAD and detection limits
5 Also appears to be near or at its theoretical limit for detection
What is the authors take on high efficiency atomization methods
Can be designed for the most specific methods such as fluorescence in mind which is good
They mostly involved thermal generation of relatively volatile elements held in quiescent environment for longer observation times
However this is specific and most elements are not volatile and require more energy (such as the furnace) to vaporize
Likes sputtering
Explain ion sputtering
1st) Sample is conductive (if not make conductive by mixing (admixing) with graphite)
2) put in chamber with inert gas (several torr pressure- high pressure)
3) Hold sample at high negative voltage (500 V( - current of 5-300 mA allowed to pass (at the cathode?)
4) this results positive ions of inert gas are generated and attracted to cathode - causing bombardment of this cathode with inert gas ions which fragments the surface (these fragments are hopefully our atoms
5) atoms now in a dense cloud above sample surface ready for SAD hopefully
Pros and Cons of sputtering
Pros: “outside in process” cloud of atoms in a gradient (less as move farther away) so can sample the various depths
2) no explosive sample fragmentation in thermal volatilization
3) Generated in a quiescent environment allows for long observation times(better sensitivity9)
4) High efficiency
CONS:
1) Slow- requires minutes to form a stable population
2) samples must be conductive
3) must be held in a closed, evacuated chamber which is the biggest issue (real samples I guess)-experimentally inconvenient
Talk about potential imrpvoements in sputtering
Atmospheric sputtering can make it quicker (micro arc technology)
- also requires no vacuum
and can use non conductive samples (just requires conductive substrate)
Author sees potential for this and MIP (microwave ICP microwave induced plasma - like ICP but worse less efficient, less dilution though ) in conjunction as promising
What are current means for atom or ion trapping and what are the authors ideals
chamber at reduced pressure Currently can do in a closed d like sputtering
OR let atoms slowly escape from a heated tube (semi enclosed) - this is what is done in furnaces -however short of SAD capabilities (not as good as enclosed as well - shorter time)
His ideal is trapping for VERY long times hours , day minutes etc - undersuch times could identify every element in a sample
What are the uthors thoughts on mass spec
Can use linear ion trap to hold atoms for longer periods of times
Know the basic 4 rods RF and DC create potential well to trap our ions - - should be able to probe with a laser
only issue - requires ionization - LOW efficiency not great for detection
Where can selectivity be appleid to in a SAD method
In generating the ions and the detection process - (and in fluorescence you have the excitation and fluorescence processes which can be made more selective
What are ways selectivity is increased in SAD methods
1
1) Using very specific narrow band dye laser - often selective enough to excite specific elements or specific isotopes, (can still be interfered with by background radiation or secondary ionization)
What is the problem of making SAD more selective on the detector end
ANy attempt here jst reduces detection efficiency (eg incorporating a mass spec) - often ions can only be analyzed once as well (fluorescence not the case) -
but generally reduces the efficiency
Believed would make it unusable for SAD
What are ways to make LEAF more selective
THe following are LEAF specific:
2)Using non resonance fluorescence (so the emission isn’t at the same wavelength at absorption - somewhat separating it (still has general issue of noise though - also considering hot environment of furnace and multiple elements fluorescence - can still have overlap
3) Can use a two photon or double resonance process in which 2 lasers are used - one is done continuously - brings it to one energy level - the other is pulsed and brings it to another high energy level - can detect time wise for this second higher emission - distinguish from other elements that might have that first energy level
4) Can do resolution on a chemical means - - separate elements -modify atmosphere to scavenge selectively unwanted elements (can also use specific discharge gases like He Ar or Ne ) to do this
5) Can do it time based either by
A) Elements volatilize at different times - can employ temperature ramps to either avoid or set it at specific times for detection
B) AToms have intrinsic excited state lifetime so can time its fluorescence -, Atoms excited by laser under go first order kinetic decay characteristic of a specific atom (also importantly scattering is much shorter time scale - see figure - can wait to just analyze fluorescne)
What three means do they use of evaluating modern day methods
1) determine the limiting noise source - determines the lowest concentration that can be measured
2) Determine the magnitude of signal expected from an atom based on its fundamental characteristics
3) Teh degree to which free atoms are generated
What is a common source of noise
Often it is photon detectors - receiving unwanted photons - often noisy -
What is intrinsic LOD
SO starts with talking about intrinsic LOD – ultimate limit for a method determined statistically – based on physical interaction of analyte atoms or ions with photons, electric or magnetic fields
What are the equations related to intrinsic LOD
2 equations one for the MEAN NUMBER OF EVENTS REGISTERED AT DETECTOR
fd * Np
Fd is detection efficiency – which relates to mean probability of generating one event per sample atom in the probe volume per second
Np is mean number of analyte atoms in probe volume
ANd
Minimum # of atoms detectable in robe volume
k^2 / (fd * R)
k is the S?N ratio at the detection limit (also statistical confidence level)
fd is still fd
and R is the number of repeated probings
What are the assumptions in calculating the minim # of atoms detectable in a probe volume
This equation is dependant on k being the S/N at detection limit AND THE PROBABILITY OF OBSERVING AN EVENT FOR ON EATOM PRESENT IN VOLUME IS SMALL
- Also presupposition that the background and detector dark current counts are negligible (which is a simplificaition)
-
What is intrinsic LOD
ultimate limit for a method determined statistically – based on physical interaction of analyte atoms or ions with photons, electric or magnetic fields
Run through the calculation for ELECTROTHERMAL AND ATOMIC ABSORPTION SPECTROMETRY (ETAAS)
So 3 equations - Absorbanced measured
minimum number of atoms detectable which is what we care about and
number density
THe key is number of atoms detectable is Amin * pi * r^2 / Qa
of which r ^2 is in meters and Qa is generally given (ranges between) 1E-19 and 5E-16
this produces 6E7 to 3E11 atoms BUT to get to a single atom
it would need to be 0.01 - 0.8 uM diameter - and would need an optical beam that size which is not possible
Familiairize self with LEAF equations and units
Npmin = k^2 / (fd * ea * ei * Tm)
fd no units
ea no units atomizer efficiency
e1 - Hz laser reptition frequency
Tm secondsmeasuring time (residence time in the furnace)
Fd - depends on a lotAtomizer volume
Illumination of volume
Collection efficiency for fluorescence radiation
The solid angle and transmission of the optical system
The quantum efficiency of the detector
Other factors related to specific atomic transitions involved
When LEAF Npmin is calculated what is the takeaway
Can achieve SAD - HOWEVER - requires continuous wave laser or pulsed laser with high repetition frequency and these have not yet been performed at the time
Review Trapping of atoms (RF trap and the details of that - hard to put into a question
Describe radiation pressure and its use in trapping
IN terms of momentum when an atom absorbs - takes the unimolecular momentum of the in coming light beam -
-the momentum changed caused is h/lp with h being planks and lp being photon wavelength
The emission resulting Goes everywhere - canceled out - no resulting momentum so all that remains is the momentum from the incoming
This can be used to influence direction of atoms or trap in a sense
PLanck constant units
Planck constant h is equal to exactly 6.626 070 15 × 10-34 Joule seconds.
What are the equations for atom beam velocity and momentum
v = (2kT/m) ^ (1/2)
where k is boltzmann and T is temp and m is atomic mass
velocity change per absorption is
dV = h/(m*lp)
Boltzmann constant units
1.380649 × 10−23 joule per kelvin (K)
What is a downside to using radiation pressure
- when atoms are deccelerated - their resonance frequency shifts (Doppler effect) - it’s detuned with respect to the frequency of the counter propagating laser beam - Need to do something to overcome this (TABLE 2)
What is optical molassess and its use for SAD
using 3 perpindicular lasers to cool and trap neutral atoms
requires lower energy so can be done with diode laser - simple tech - can trap atoms for 0.2 seconds at temp of 100 uK
can have some issues eg need deceleration length of 1 M
OPTICAL trapping for SAD setup
See document - basically shoot it intersecting the atom beam so some of them take a different trajectory where they are then detected with a suitable laser beam using fluorescence signal for LEAFS detection (note this is OPTICAL COOLING not optical molasses)
What are advantages ICP - MS introduced
simpler spectral background, higher sensitivity and completel elemental coverage and isotpic resolution
Go through equations for ICP-MS
Using teh efficiencies of each part of the MS - one can get an overall detection probability for ICP-MS
P det = 3E-11 - 1 E-9 counts per atom
So using equation 2 (for NP(min)) the overall LOD becomes
(Namin)^8 - 3E8 atoms - read article
dependant on operating in SIM mode with measuring time of 10s
Downsides to ICP MS for SAD
LOD increases proportionally to number of elements
Also depending on parts used (eg ionization source - can adjust LOD
at current point in time - WAY too inefficient for SAD capability - efficiency too low at sample interface and throughout the quadrupole (will not reach below 1E6 atoms
What makes the TOFMs attractive and how does it interface with current SAD attempts
1 more efficient than triple quad
2 Can volatilize our sample under vacuum with TOFMS (no need for the multi step pressure change - again loss of ions)
Can use electrothermal volatilization
3While the TOF needs to be pulsed can use an ATOMIC BEAM TOF - AB TOFMS which has an ion source trap to pulse ions into the TOF
aLSO IS tof - HIGHER RES (at time > 1000)
Mechanically moderate compared to QQQ
requires fast data processing (fast instrument 27 kHz frequency of operation
review tof diagrams AND COMPARE EFFICEINCIES TO QUADRUPOLE
wHAT IS THE tof EQUATION
t FLIGHT = L * (2 /Ua)^ (1/2) * (m/z) ^(1/2)
With t flight being time of flight
l being flight length
Ua being acceleration voltage
and m/z is m/z
Where is ABTOFMS at the time of the paper
not far from sad capability with isotpic reoslution
What does practical AB TOFMS look like
Used a tungsten toil for volatilizer - used a reflection and an MCP for detection
CONCLUSIONS FROM METHOD OF DETECTION APPER
LEAFS should be capable of achieving SAD it optimum excitation lasers become available
Atom trapping and laser interaction can theoretically do SAD
Elemental MS should be able to do SAD but isn’t there currently but lots of potential (increase efficiency, of sample inlet or mass filters, TOF is promising like it -
ALSO MS really promising due to its full elemental range, multi element capacity, expect MS to be much broader
ABSOLUTE LOD vs RELATIVE LOD
absolute is the minimum amount of analyte detectable
relative is the minimum concentration
Does the paper indicate there is a right way to report ta detection limiti
No as long as it is clear what it is referring to analytically - As long as the data is interpreted correctly - and what you are writing is linked to a specific analytical task
What is the difference in developing an analytical methodology on a novel principle vs developing a detection procedure and what point does it emphaszie
In developing a new methdology - the goal is to tune instrumental parameters for best S/N
whereas a new detection procedure - care about sample amount available precision , absence of systematic error etc
THe key here is USING THE RIGHT TERMS TO STRESS the goals of your method and emphasize its benefits
Give an eample of the difference between the two
Method using RIMS for Pd and Rh in sea water - really good sensitivity however the LODs are for concentration and INCLUDE THE WHOLE METHOD - including sample concentration and separation (so good relative LOD)
the actual absolute LOD is not very impressive compared to other methods
Another example - they say they have a method for SAD with RIS - this refers to simply being able to see one atom - absolute - NOT relative - cannot see one atom in a sample
What are some key poitns in comparing methods
So again relative vs absolute - one can have much better absolute but not be great at relative
Keeping parameters the same when comparing
a major one being maybe sample volume (eg HR ICP MS can consume several mL and have a giant integration time allowing for low [] detection whereas another method can use a few micro liters have a better absolute but worse relative LOD
Instrumental limitations in amount of sample that can be analyzed is key for thinking between the two absolute and relative
What is the theoretical limit for applied analysis
if random sampling defined by square root law of particle counting which represents the lower limit for sampling error. The counting error for quantitative determination will be 1% or less if 10,000 atoms counted which gives us a minimum mass to look at - I guess the idea is that below that your error goes up and analysis is less sure
Be familiar with some unit conversion sfemto, atoo etc y
femto is 10-15; 1 ag,10~18 g; 1 zg,10~21 g; 1 yg,10~24 g
Ultratrace vs micro analysis
ultra trace should use relative LOD, microanalysis for Absolute
What are the main conclusions from omenettos paper
1) absolute LOD is still great and can be relevant
2) should not stress absolute LOD when there are no limits to sample size available for analysis
3)For absolute LOD - consider EMP - electron microprobe, Auger Electron Spectormetry (AES) and SIMS (ag level sensitivity)
4) Detection limit should be reported in a way relative to the task at hand (or both)
5) The method of comparison again depends on the task at hand again as to what parameters are key to compare
What is the main takeaway from consideration on concentration units
On the X axis for concentration for Mole fraction not weight fraction - given equation 11 weight fraction doesn’t equal mole fraction unless the molar masses of the nalayte an matrix are equal and really it comes down to the emission signal relates directly to the atoms and the moles and as such the mole fraction ins what matters not the weight fraction and more importantly they differ
What are the 3 scenarios they give in the concentration consideration papers
X in A - so plotting weight fraction and mole fraction will show the difference between weight % and mole fraction but will ultimately show the same effect eventually = - self absorption so need to look at more interesting scenarios to show thecharm
X in A vs X in B - At the same weight %
If we see a differing signal when plotted one might be tempted to say matrix effects BUT (assuming weight % in solid translate to that in plasma) unless the molar masses of matrix A and B are the same - they’re going to have different plots just given by our equation so can’t chalk up and difference in plots to matrix effects. In this scenario when you plot mole fraction you get a unified plot showing no matrix effects
The opposite case - so we have x in matrix a and b and AGAIN the same weight % But this time - the signal on the weight % plot is the same. The assumption would be NO matrix effect. However, again, if its the same signal - it should be the same mole fraction which we know cannot be true 9same weight % and mole% ) unless Molar mass A and molar mass B are equal
What is equation 11 from concentraiton
Xa = (wx * Rax)/(1+wx(Rax-1)) with wax being the weight fraction xa being mole fraction and Rax being the ratio of molar masses
What is the most important quality for standard reference material for small sample analysis>
HOMOGENEITY
How does Danzer define analytically homogenous solids and what is a key takeaway from this definition
If fluctuations in the chemical compositions determines across the sample volumeare not significantly larger than the error of the analytical procedure
THIS MEANS HOMOGENITY IS NOT AN ABSOLUTE MATERIAL PROPERTY BUT RELATED TO THE ANALYTICAL METHOD
What is the critical mass in regards to sample homogeneity
basically as the mass goes down sample heterogeneity becomes more and more important - above this critical mass Is only where we can see the difference between homogenous and htereogneous - see the image (figure 6 and explain) I think it’s like at low masses - the variance due to sample heterogeneity is so big that can’t tell the difference between heterogenous and homogenous but at higher masses - the variance from the heterogeneity plays less of a role and the homogenous eventually reaches the point where all the variance is just from the method whereas a heterogenous sample will reach a point below the variance from the method and m crit is where this divide is
OK so at small masses - the heterogeneity plays a big role in variance as the mass gets bigger it plays a smaller role. At the critical mass is where their contribution to the total variance differs whereas the heterogeneity one approaches 0 the homogenous one approaches just the s from the method - so I guess if the variation stays at the level of the method so the contribution from heterogeneity can be seen better as its just whats beyond variance beyond the method variation (so while the variance due to heterogeneity goes down - it becomes easier to assess past the critical mass) and more importantly - BELOW this critical mass - you cant tell if hetero or homogenous because the variance is just so much higher than s method for both homo and heterogenous
Critical mass graph
OK so I think this graph is saying what % does each part of the sample contribute to the total variance
Basically at low amounts it’s the same but as we increase mass – heterogenous goes down to being 0 homogenous just goes down to the s method amount
List some eamples from the Omenetto paper - Microtrace vs Ultra trace
His example was having a sample with the same concentration but the sample amount kept decreasing - the ultratrace analysis (or relative LOD) remained the same while the absolute or microtrace had to be pushed farther
Instrumental example from Omenetto Cume
compared HR-ICP-MS to GF-LIF graphite furnace Laser induced Fluorescence - the idea being that ICP MS can use mL of sample while GF-LIF only uL despite having better absolute LOD
Looking at thallim in water
Absorption and emission energy diagrams
Absorption is just a straight arrow up and emission is a wavy arrow up and a straight arrow down
