Recommended
Recommended
More Related Content
What's hot
What's hot (20)
Similar to Introduction to UV-based detectors
Similar to Introduction to UV-based detectors (20)
More from Oskari Aro
More from Oskari Aro (9)
Recently uploaded
Recently uploaded (20)
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
- 12. 12 pH Effects Anthocyanin pigment in buffers of varying pH
- 13. 13 pH Effects Anthocyanin pigment in buffers of varying pH
- 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
- 29. 29 Flow Cells – Thermo Scientific™ UltiMate™ 3000 HPLC System
- 30. 30 Inlet Outlet Light Flow path Flow cell locks Glass body Flow cell FluidicsFiber optics Vanquish Light Pipe Flow Cells – Fused Silica
- 31. 31 • Overlay of chromatograms obtained from different test setups • VDAD: higher (by 30-50%) and narrower peaks (by 30-35%) Column: Thermo Scientific Hypersil Gold, C18, 2.1 ×100 mm, 1.9 µm, P/N 25002- 102130 System: Binary UltiMate 3000 Mixer Vol.: 200 µL Mobile Phase: A – Water B – ACN Flow rate: 0.7 mL/min Pressure: 630 bar (max) Temperature: 35 ºC Injection: 1 µL Detection: DAD-3000RS, semi-micro or analytical flow cell, wide slit (4 nm) Vanquish DAD with standard flow cell, 4 nm slit 254 nm, 4 nm bandwidth, 20 Hz, 0.2 s response time Analytes: 1. Uracil 2. Acetanilide 3. – 10. Homologous Phenones 50 µg/mL each 0.00 3.50 -50 0 500 ACN: 40 % 100 % 40 % 1 2 3 4 5 6 7 8 9 10 min mAU 0.30 0.40 -30 0 250 1 min mAU Vanquish DAD, 2 µl standard flow cell DAD-3000RS, 2,5 µl semi-micro flow cell DAD-3000RS, 13 µl analytical flow cell Binary UHPLC Gradient Performance
- 32. 32 • VWD • 11µl 10mm Standard Analytical flow cell • 2,5µl 5mm Semi-Micro flow cell • 45nl 10mm Capillary flow cell • 3nl 10mm Nano flow cell Flow Cells • DAD • 13µl 10mm Standard Analytical flow cell • 5µl 7mm Semi-Analytical flow cell • 2,5µl 5mm Semi-Micro flow cell • Vanquish • 2µl 10mm Standard Analytical Light Pipe flow cell • 13µl 60mm Light Pipe 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)
- 35. 35 Diode Array Detector Slit width • Slit defines optical resolution and therefore minimal physically meaningful bandwidth Flow Cell
- 36. 36 Diode Array Detector Bandwidth S/N ratio Spectral resolution ↑ ↑ ↓ ↓ ↓ ↑ Slit width Baseline noise Spectral resolution ↓ ↑ ↑ ↑ ↓ ↓ Slit width Bandwidth Flow Cell • Slit defines optical resolution and therefore minimal physically meaningful bandwidth
- 37. 37 Effects of Slit Width Slit width - 1nm 4 x light > 0.5 x noise
- 38. 38 Effects of Slit Width Slit width - 16nm 4 x light > 0.5 x noise
- 39. 39 Setting Bandwidth Bandwidth 4 nm s/n = 5 4 x light = < 0.5 noise
- 40. 40 Setting Bandwidth Bandwidth 4 nm s/n = 5 Bandwidth 30 nm s/n = 25 4 x light = < 0.5 noise
- 41. 41 Linearity of butylparabene • Measured at 240 nm • The linearity is excellent between 1-4 nm bandwidth • From 8 -100 nm the linearity permanently decreases 0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 0.99 0.991 0.992 0.993 0.994 0.995 0.996 0.997 0.998 0.999 1 1 nm 2 nm 4 nm 8 nm 16 nm 32 nm 64 nm 100 nm Bandwidth Coeff. Of Determination Linearity RSD 190 250 300 350 402 -10.0 0.0 10.0 20.0 30.0 40.0 50.0 60.0 194.1 257.6 nm % Butylparabene Influence on Linearity – Bandwidth
- 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)
- 55. 55 Stray Light 0.00% Stray light 0.01% Stray light 0.10% Stray light 1.00% Stray light True absorbance (AU] Measuredabsorbance[AU] Primary effect is to reduce analyte absorbance / Sensitivity
- 56. 56 Noise: Definition • Short term (Statistical) signal changes • Defined in ASTM • Defines LOD and detection accuracy 0,0 1,0 2,0 3,0 4,0 5,0 6,0 7,0 8,0 9,0 10,0 11,0 12,0 13,0 14,0 15,0 16,0 17,0 18,0 19,0 20,0 21,0 -0,0150 -0,0125 -0,0100 -0,0075 -0,0050 -0,0025 0,0000 0,0025 0,0050 0,0075 0,0100 0,0125 0,0150 0,0175 0,0200 8000927 #24 UV_single_dry_noise_drift UV_VIS_1 mAU min WVL:254 nm Ratio to Ambient_Temp 23,32 23,60 23,80 24,00 24,20 24,40 24,60 24,80 25,00 25,20 25,40 25,60 25,80 26,00 26,20 26,40 26,60 26,80 27,00 27,20 27,40 27,70 -0,26 -0,20 -0,10 -0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 System03#6 [modified bygyhaa] Test 5PGM: QS_Cyto_C_NAN_WPS_T UV_VIS_1 mAU min WVL:214 nm
- 57. 57 Noise • Effects measurements on analytes with low absorbance • Also affects high absorbance samples 0,0 1,0 2,0 3,0 4,0 5,0 6,0 7,0 8,0 9,0 10,0 11,0 12,0 13,0 14,0 15,0 16,0 17,0 18,0 19,0 20,0 21,0 0,0000 0,0100 0,0200 0,0300 0,0400 0,0500 0,0600 0,0700 0,0800 0,0900 0,1000 ND_dry_turnon #9 1 UV_VIS_1 min Ratio to Electronics_Temp
- 58. 58 Noise Usable linear range 0.1 – 1 AU Stray light error Stray light + noise error curve Stray light –noise error curve %Error Absorbance (AU]
- 59. 59 Webinar: Introduction to UV-based Detection Thank You! Any Questions?
Editor's Notes
- 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