08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
Introduction to UV-based detectors
1. The world leader in serving science
Jan Pettersson
Nordic HPLC & Chromeleon CDS Support Specialist
Thermo Fisher Scientific
Introduction to UV-based Detection
2. 2
UV Vis Detectors
• The ideal detector ?
• Why do we get a signal?
• Optics
• Lamps
• Flow cell
• Band and slit width
• Data collection rate and time constant
• Reference
• Stray light, refractive index effects & noise
Thermo Scientific™ Vanquish™ HPLC
3. 3
UV Vis Detectors – The ideal detector?
• A workhorse for detection and quantitation of organic compounds
• Fast, accurate and sensitive
• Quick and easy to set up and require only basic maintenance
• However – to optimize performance one needs to understand the
basic physics of the absorption process and the working principles of
the various instrument designs
• This will lead to improved sensitivity, robustness and quantitation.
MWD DADVWD
4. 4
UV Vis Detectors
• The ideal detector ?
• Why do we get a signal?
• Optics
• Lamps
• Flow cell
• Band and slit width
• Data collection rate and time
constant
• Reference
• Stray light, refractive index effects & noise
5. 5
Why Do We Get a Signal?
Very simplified:
• The light from the lamp excites the electrons in the sample to
a higher state of energy.
• Different molecules absorbs light at different frequencies.
• The shorter the wavelength the higher the energy
6. 6
• Structures with double bonds often absorb UV light due to the
presence of p electrons
• Hetero atoms having non-bonding valence-shell electron
pairs also absorb UV light
• Conjugated systems are particularly useful in UV
spectrometry
Why do we get a signal?
7. 7
Above 200 nm excitation of electrons from p-, d-, and π-orbitals and especially
conjugated systems produce informative spectra
UV Chromofores
8. 8
UV spectra often contain broad responses due to vibrational and
rotational transitions which accompany electronic transitions of
bonding electrons
9. 9
• Series of alternating single and double or triple bonds represent a conjugated
system
• Addition of double bonds increases the wavelength by around 30nm
• Intensity of
absorption increases
due to delocalized p
electron systems
giving rise to a larger
number of ‘pseudo’
double bonds
Conjugation effects
10. 10
Solvent Effects – Polarity
UV-Visible spectrum of
propanone in hexane
UV-Visible spectrum of
propanone in water
11. 11
Solvent Effects – Why Water and ACN are Popular
Solvent (nm) Minimum wavelength
Acetonitrile 190
Water 191
Cyclohexane 195
Hexane 201
Methanol 203
Ethanol 204
Ethoxyethane 215
Dichloromethane 220
Trichloromethane 237
Tetrachloromethane 257
14. 14
• Expansion of the solvent, may be sufficient to change the apparent
absorbance and thereby the accuracy of quantitative results
• Temperature may affect equilibria, which can be either chemical or
physical (denaturation of nucleic acids for example)
• For organic solvents, changes in refractive index with temperature can
be significant
• Convection currents cause different temperatures to occur in different
parts of the cell, the resulting Schlieren effect can change the apparent
absorbance.
• To control this problems have as stable temperature as possible!
Temperature effects
15. 15
UV Vis Detectors
• The Ideal detector ?
• Why do we get a signal?
• Optics
• Lamps
• Flow cell
• Band and slit width
• Data collection rate and time constant
• Reference
• Stray light, refractive index effects & noise
16. 16
• Two UV/Vis detectors
• Forward optics design
=> variable wavelength detector (VWD)
=> no real-time spectra information available
• Reversed optics design
=> diode array detector (DAD)
=> includes real-time spectra information
Operating Principle – Optics Design
17. 17
Operating Principle: Variable Wavelength Detector (VWD)
Forward optics design
• Only the selected wavelength
passes the flow cell
• A part of the light beam is
redirected to the reference diode
Light source Dispersion device Flow cell Sample diode
18. 18
Operating Principle: Wavelength Diode Array Detctor (DAD)
Reversed optics design
• Light beam passes the flow cell before being diffracted
• No true reference signal can be obtained
• Any diode or bunch of diodes can be selected as a reference
Light source Flow cell Dispersion device Diode array
19. 19
Uses of Diode Array Detectors
Peak purity
measurement
Signal deconvolution /
Multiple wavelength
acquisition
Dynamic spectral acquisition
and identification
254 nm
220 nm
20. 20
Operating Principle – 2nd Order Filter
• Light diffraction at a dispersion device results always in different orders of light
segmentation
No filter
Dispersion device
(grating or prism)
‘White’ light
0-order
1st-order
2nd-order
21. 21
Source / UV Lamp
• Deuterium lamp – silica quartz envelope to reduce noise and transmit at lower
wavelengths mounted on a pre-set (focused) mount
• Need a stable source with adequate intensity over a suitable wavelength range
• Over time the intensity decreases – 1000h half life is typical
• It loose about 6 hours every time it is switched on and off
22. 22
UV Vis Detectors
• The Ideal detector ?
• Why do we get a signal?
• Optics
• Lamps
• Flow Cell
• Band and Slit Widht
• Data Collection Rate and Time Constant
• Reference
• Stray light,Refractive Index Effects & Noise
23. 23
Source / Vis Lamp
• Tungsten Halogen lamp for the longer wavelengths
• Light from both sources can be mixed to generate a single broad band
source
24. 24
• ‘UVLampIntensity‘: measured at 254 nm
• The reading is normally above 3 million counts (narrow slit width) for the
UV lamp for an analytical SST cell, and should be above 1 million counts at
least
• ‘VisLampIntensity‘:
• Expected value: above 2 million counts
• Beside intensity value noise level and
signal-to-noise ratio need to be checked
(application depending)
Guidance values:
VWD: OQ limit 50%, PQ Limit 40%
DAD: > 3.000.000 counts (depending on slit width, here: narrow)
Lamp behavior Intensity
high at the beginning of lifetime
unstable and highly drift and noise affected
Significant intensity drop within the first ~ 100 hours
afterwards slowed and steady decrease
Source / Vis Lamp
25. 25
UV Vis Detectors
• The Ideal detector ?
• Why do we get a signal?
• Optics
• Lamps
• Flow Cell
• Band and Slit Widht
• Data Collection Rate and Time Constant
• Reference
• Stray light,Refractive Index Effects & Noise
26. 26
Flow Cell – Signal to Noise
Signal height
• Light path should be as long as possible
• Small flow cell volume, small inner diameter measurement channel
SampleLamp
Detector
Flow inFlow out
27. 27
Flow Cell – Signal to Noise
Signal height
• Light path should be as long as possible
• Small flow cell volume, small inner diameter measurement channel
SampleLamp
Detector
Flow inFlow out
To reduce Noise
• Get as much light as possible through flow cell
• 4 x light > 0.5 x noise
28. 28
Flow Cell Volume
• Flow cell volume should not exceed 10% of the peak volume
Flow cell ok
PeakVolume
Cellvolume
Flow
Flow cell too big
Micro column
peak volumes
Flow cell ok
Micro column
peak volumes
Smaller cell volume less light is passing through the flow cell
33. 33
UV Vis Detectors
• The Ideal detector ?
• Why do we get a signal?
• Optics
• Lamps
• Flow Cell
• Band and Slit Widht
• Data Collection Rate and Time Constant
• Reference
• Stray light,Refractive Index Effects & Noise
34. 34
Operating Principle – Slit Width
• VWD detector
• Bandwidth is defined by entrance slit
• At 254nm the bandwidth is 6nm (254nm +/- 3nm)
42. 42
• Linearity of butylparabene
• Narrow slit always provides both a slightly better correlation coefficient
and a slightly better linearity RSD
0%
1%
2%
3%
4%
5%
6%
7%
8%
1 nm 2 nm 4 nm 8 nm 16 nm 32 nm 64 nm 100 nm
LinearityRSD
Bandwidth
Narrow Slit Wide Slit
Influence on Linearity –Slit width
43. 43
Recommended Parameters: Data Acquisition
250 250
0,100 0,105 0,110 0,115 0,120
0
50
100
150
200
Absorbance[mAU]
t [min]
0,100 0,105 0,110 0,115 0,120
0
50
100
150
200
Data rate: 5 Hz
t [min]
Absorbance[mAU]
Data rate: 100 Hz
• Too few data points effect peak form, reproducibility and area precision
• A minimum of 20, ideally 30-40 data points/peak is required
45. 45
Time Constant
• The rise time (Response time) is closely releated to the time
constant:Rise time = 2,2 x Time constant
46. 46
Data Collection Rate and Time Constant
• Noise is much more influenced by time constant than by data collection
rate
-0,0650
0,0200
1 - Det_Set_Noise #1 DCR fix at 25 Hz and TC varied UV_VIS_1
mAU
1
WVL:245 nm
0,0 2,0 4,0 6,0 8,0 10,0 12,0 14,0 16,0 18,0 20,0 22,0 25,0
-0,0850
-0,0500
0,0000
2 - Det_Set_Noise #3 TC fix at 0.06 sec. and DCR varied UV_VIS_1
mAU
min
2
WVL:245 nm
TC 2.0 s TC 0.2 s TC 0.06 s TC 0.03 s TC 0.01 s
-0,0650
0,0200
1 - Det_Set_Noise #1 DCR fix at 25 Hz and TC varied UV_VIS_1
mAU
1
WVL:245 nm
0,0 2,0 4,0 6,0 8,0 10,0 12,0 14,0 16,0 18,0 20,0 22,0 25,0
-0,0850
-0,0500
0,0000
2 - Det_Set_Noise #3 TC fix at 0.06 sec. and DCR varied UV_VIS_1
mAU
min
2
WVL:245 nm
DCR 1 Hz DCR 10 Hz DCR 25 Hz DCR 50 Hz DCR 100 Hz
47. 47
Sampling and Rise Time
• The same instrument, back pressure loop, eluent and sample. The area is
the same – the peakshape is very different.
• 2,5 Hz 2s response time
• 10 Hz 0,5s response time
48. 48
Automated Settings
• The Program Wizard of Thermo Scientific Chromeleon™ 7.2 CDS has
a dedicated step for setting the correct ‘Data Collection Rate’ and ‘Time
Constant’
• The internal calculation is based on the peak width at half peak height
of the slimmest peak in the chromatogram
49. 49
UV Vis Detectors
• The Ideal detector ?
• Why do we get a signal?
• Optics
• Lamps
• Flow Cell
• Band and Slit Widht
• Data Collection Rate and Time Constant
• Reference
• Stray light,Refractive Index Effects & Noise
50. 50
Operating Principle – Reference on a DAD
• Light beam passes the flow cell before diffracted
No true reference signal can be obtained
• Any diode or bunch diodes can be selected as a reference
• If selected reference and acquisition wavelength are the same, the
resulting signal will be zero (0)
• Select a reference wavelength in a ‘quiet’ area of the spectrum
(where little absorbance occurs)
• Reference wavelength range
should not interfere with
absorbance range of any
compound of interest
51. 51
Operating Principle – Reference on a DAD
• A reference can compensate for
• Fluctuations in lamp intensity
• Changes in absorbance/refractive index during gradient analysis
• Background noise
52. 52
Issues with Reference Wavelength
Blue Chromatogram: Without reference Black Chromatogram: With reference
• Wavelength: 254 nm
• Both the UV and Vis lamps turned on
• Reference wavelength set to 600 nm (80 nm bandwidth)
53. 53
UV Vis Detectors
• The Ideal detector ?
• Why do we get a signal?
• Optics
• Lamps
• Flow Cell
• Band and Slit Widht
• Data Collection Rate and Time Constant
• Reference
• Stray light,Refractive Index Effects & Noise
54. 54
Stray Light
• Stray light is radiation emerging from the monochromator of all wavelengths
other than the bandwidth at the selected wavelength
• Arise from imperfections in the grating, optical surfaces, diffraction effects as
well as wider bandwidth and slit width settings
𝑨 = 𝜺𝒍𝒄
ε = Molar absorption coefficient (dm3 mol-1 cm-1)
l = Path length (cm)
c = Concentration (mol dm-3)
The presentation today is an in depth presentation of UV/Vis detectors
I guess that you have heard almost all of this at university, but it might be as it is for me – it was 30 years ago and I don’t remember all – if any of what they tried to tell me. Hopefully I have compensated what I have forgotten by learning things in stead. This presentation is based on 30 years of experience of these detectors and I hope that it will be useful in your daily work.
The UV/Vis detector is the most common detector for HPLC.
Could the reason be that it’s the ideal detector for HPLC?
It’s a workhorse for detection and quantitation of organic compounds
It’s Fast, accurate and sensitive
Quick and easy to set up and require only basic maintenance
However – to optimize performance and monitor critical functions one needs to understand the basic physics of the absorption process and the working principles of the various instrument designs.
This will lead to improved sensitivity, robustness and better quantitation
The first part is why do we get a signal?
This is a short version of a university lesson I appologise if you already know all of this
Very simplified:
The light from the lamp excites the electrons in the sample to a higher state of energy. This make the sample to absorb the light.
Different molecules absorbs light at different frequencies.
The shorter the wavelength the higher the energy.
Above 200 nm excitation of electrons from p-, d-, and π-orbitals and especially conjugated systems produce informative spectra.
This is a list of what wavelengths different bonds absorbs.
Unfortunately it is not straight forward because most molecules conatins several different bonds that absorbs light in several different ways.
This is one of the reasons you will get a spectra and not just a line.
However by using the spectra you can identify what molecule you have detected, or at least get a good idea.
This is a conjugated system, a series of alternating single and double or triple bonds
Addition of one double bonds increases the wavelength by around 30nm because the electrons are easier to excite
Here is an example when the solvent is changed from propanone in water to propanone in hexane.
This is also one of the reasons you can’t buy good libraries of UV spectras as you can do with GCMS. They UV spectras differ a lot depending on the eluent.
Everything influence the spectra, it’s not only the molecule it self but also other things like solvent, polarity, pH and temperature
It this example it’s polarity if the solvent. Polar solvents gives longer wavelengths
The excited states of most transitions are more polar than their ground states because a greater charge separation in the excited state.
With polar solvents the dipole–dipole interaction reduces the energy of the excited state more than the ground state, hence the absorption in a polar solvent will be at a longer wavelength. As you remember light with longer wavelength contains less energy
The reverse is also possible if the excited state reduces the degree of hydrogen bonding. Here the transitions are n→π* and the shift of wavelength is due to the reduction in solvent hydrogen bonding
To make it even more complicated the molar absorption constant also changes from solvent to solvent
Have you ever wondered why Acetonitrile is such a popular solvent for HPLC?
The reason is that it absorbs light at a low wavelenght and hopefully don’t absorbs at the same wavelenght as the sample.
You will not get any information out of your sample is the eluent absorbs all the light.
A change in pH usually change retention times
It might also effects absorbance. This can be very large if you are unlucky and the equilibrium shifts between two different forms
In this diagram you can see really big differences when the pH is changes. The same sample is run at pH 2 , 6 and 10
If you decide to use a buffer to control this, remember that they can also absorb light
The example is a bit extreme, it is anthocyanin and that changes between red, purple, or blue depending on pH
You can find it in food plants such as blueberry, raspberry, black rice, and black soybean
Now we are thrue the basic university training and we are entering the techical part.
To know how to handle a detector you need to know who it works
There are two main detector designs, forward optics that are used for variable wavelength detecors and reversed optics that are used for Diode Array Detectors
In a detector with forward optics the light from the lamps first goes thrue a slit, then it is divided into different wavelenghts with the help of a grating or a prism.
Then a small portion of the light is diverted by a mirror to be used as reference and the rest of light hits the flowcell
In this case we have a true reference that will be used to calculate how much of the light that was absorbed by the sample
It a detector with reversed optics all of the light goes thrue the flow cell before it goes thrue the slit and then is divided by the prism or the grating.
All frequences are aquired at the same time by a diode array and you will get a spectra telling what frequences of light that are absorbed by the sample which is very nice and give extra information
This is examples of what you can use a Diode Array Detector for
Peak Purity Measurement, to see if you have the same spectra over the peak. If it changes between the start and the end of the peak you have two different compounds coo eluting.
Signal deconvolution, in this case you acquire two or more wavelengths at the same time to get better signal to noise ratio and also less problem if the peaks are coo eluting because you are using different wavelengths for the different peaks
Spectral acquisition can be used to help identifying peaks.
When you split up the light you will get something like a rainbow that send the light to a diode array that measure the intensity of the different wavelengths.
Have you noticed that when you look at a rainbow there are actually one more rainbow with less light close to it?
This is 2nd order segmentation and also happens in a detector
0-order is ‘white’ light from the lamp with no segmentation
1st order is segmentation of the white light into the different spectral components
2nd order this is the same segmentation as on 1st order but on a different place of the diode array and it’s less intense
As a consequence the acquired data includes the selected wavelength, but also light of twice the selected wavelength and 3-times and also 4-times the wavelength
The deuterium lamp that is used for UV has a very high intensity when it is new and it’s also unstable, drifts and is noisy
The Intensity significantly drop within the first ~ 100 hours afterwards slowed and steady decrease and the light is stable
A lamp typically loose about 50% of the intensity in 1000 hours
It also loose about 6 hours every time you switch it off and on again, sometimes it is better to leave it on if you run lots of samples every day
To get light in higher wavelenghts where an UV lamp is not efficient some detectors have a tungsten lamp
This lamp has a long life time usually more than 10.000 hours
Lights from both lamps are mixed together to get a broader range
When you look at the different intensities at the different wavelength it is not linear at all.
To compensate for this the detectors have software that make the intensity more even for the different wavelengths
To get a good signal the light path should be as long as possible
In combination with a small flow cell volume, small inner diameter measurement channel the detector is going to have a longer part of the cell filled with the sample
Half the inner diameter will give four times as high signal
To reduce Noise
Get as much light as possible through flow cell
4 times the light is the same as half the noise
To decide what cell to choose is difficult becase you want a long narrow cell for sensitivity and this gives scattering effects.
A narrow cell also makes it difficult to get enough light thrue it
The size of the flow cell depends on the peak volume and the volume of the cell should not be more than 10% of the peak volume.
The peak volume is calculated as the flow times the basline width of the peak
In the example on the left the cell volume is OK as only parts of the peaks is in the cell at the same time
In the middle example the volume of the cell is so high that two peaks can be in the cell at the same time. If this would have been a real sample you would only have seen one peak with a strange peak shape
In the third example a smaller cell is used an the peaks are separated
This is what a flow cell looks like in the Thermo Scientific Ultimate 3000 detectors
They can easily be exchanged and when you put in a new cell it ”clicks” into place
The cell type is automatically recognised by the detector
It might be a good idea to have a few cells with different volumes to choose from depending on the application
On a Variable Wawelenght Detector the bandwidth is defined by the entrance slit.
The optical resolution is defined by this and you can’t get better resolution than the slit witdh on all kinds of optical detectors
If you need better resolution you must choose a more narrow slit
On the Thermo Scientific Ulitmate 3000 Variable Wawelenght Detector the slit is 6nm
This is a typical Diode Array Detector
The light will first go thru a slit that will define the optical resolution of the detector before entering the cell
After the cell the light is divided into a spectrum and collected by the diode array theese usually have 1024 diodes.
Each of them get about 2nm of the light and the bandwidth is how many of theese you bounce together.
If you as in the example have a 4nm slit there is no use to have a bandwidt less than 4nm the resolution can’t be better than the slit width anyway
When you decrese the slith width the basline noise and spectral resolution will increase
If you increase the bandwidth the signal to noise ratio will increase and the spectral resolution will decrease.
With a slith width of 1nm you get peak spectrum with high resolution, but quite a lot of baseline noise.
This setting is useful when you want to do spectra libraries
With a slith width of 16nm you don’t get as much infomation of the peak spectrum, but much less baseline noise.
This kind of setting is used when you want to quantify peaks and not necessarly need the spectrum for identification
As shown in theese examples with low bandwidth you get more noise but better spectral resolution
When the bandwidth is increased the spectral resolution decrease but you get better signal to noise values.
The relation between light and noise is four times the light is half the noise.
You need to figure out what is most important for you as high spectra information as possible or the best signal to noise ratio it is not possible to get both at the same time.
First is the data acqusition rate, this is how many data points are collected in a second
Too few data points effect peak form, reproducibility and area precision as in the left picture
A minimum of 20, ideally 30-40 data points/peak is required
Unfortunately as high data collection rate as possible is not good because higher data collection rates also increases the noise
Here is another example of the same sample run twice with different collecrion rate, it’s easy to decide what looks best.
25 times hight data collection rate is not the same as 25 times as good data
The other important parameter for data collection is the time constant.
This parameter tells how quickly the detector responds to a change in signal and the definition is how long time it takes for a peak to go from 10% to 90% of it’s final value.
Too small time constant gives a noisy baseline and too long distorted peaks.
A you see Noise is much more influenced by Time Constant than by Data Collection rate
To get good chromatograms you need to set this right and it is by far them most common problem when the chromatograms look noisy
Too low Time Constant make noisy chromatograms and too high distorted peaks
This is what can happen if you use completely wrong data collection rate and rise time.
It is the she same instrument, eluent and sample. The area is the same – the peakshape is very different.
The flat broad peak has 2,5 Hz as data collection rate and response time of 2 seconds
In the blue chromatogram the data collection rate is increases to 10 Hz and the response time lowered to 0,5s.
There are really a big difference in the peakshapes if you use the wrong settings.
Reference wavelengths works very well with variable wavelength detectors.
This is because the light is diffracted before it enter the flow cell
This is not the case for a Diode Array Detector, Here you can use any diode or bunch diodes as a reference, this might seems as a good an flexible solution.
Unfortunately all light has already been thru the flow cell and might have been absorbed by the sample or eluent
If selected reference and acquisition wavelength are the same, the resulting signal will be zero
What you should do is select a reference wavelength in an area of the spectrum where there are as little absorbance as possible
And off course the reference wavelength range should not interfere with absorbance range of any compound of interest
When it works with a reference it can compensate for
Fluctuations in lamp intensity
Changes in absorbance/refractive index during gradient analysis
Or Background noise
This is what can happen when you use a reference wavelength with a Diode Array Detector
You can get random noise, spikes or drifting baseline and several other unwanted surprises like negative peaks
Try to avoid reference wavelength with DAD
Stray light is radiation emerging from the monochromator of all wavelengths other than the bandwidth at the selected wavelength
Arise from imperfections in the grating, optical surfaces, diffraction effects as well as wider bandwidth and slit width settings
This is getting worse when a detector is getting old because of accumulated dirt and worn out mirrors in the detector
This always causes negative deviations from Beers Law – which in this case is loss of linearity
This is how linearity is affected by stray light
Typically an older detector can not be used with as high absorbances as a new detector
You need to check this at least once a year to know the limitations of your detector
This is the definition of noise
Usually 20 1 minute segments are measured and an average is calculated
What kind of problems can noise cause?
Except for the obvious problem that high noise give bad detection limits it also cause linearity problems
There are noise coming from almost everything in the detector
Photon (shot) noise arises from variation in energy emitted by a light source
The optics in the detector
The electronics
A/D converters
Temperature variations
Variances and jumps of the lamp intensity
Absorbing gases in the optics
Airfluctuations in the optics caused by temperature differences
Reftactive index variances of eluents in the flow cell
Mechanical vibrations of optical components
The list is almost endless
It affects measurements on analytes with low absorbance because the signal to noise is reduced
High absorbing samples are affected because the linearity is affected
As you can see in the examle with increasing noise and stray light you get problems with both low and high absorbing samples
You definitely need to do an performance qualification (PQ) of your system now and then to know how it performs
I hope that you ahve got a few interesting things out of the presentation and not get too scared to use an UV detector after hearing me talking about all problems. As I said in the beginning if you know how it works and the limitations you will get much better results out of your instrument