Cary Eclipse bench manual

0 · Before you start

Fluorescence is a relative measurement. Nothing the Eclipse reports is an absolute quantity, and changing a slit or the detector voltage changes every number on the screen. Read this section before you touch the settings.

What that means in practice

• Intensity is in arbitrary units. Comparing two spectra is only meaningful if slits, PMT voltage, averaging time and scan rate are identical. Record them in your notebook every time.
• Signal scales steeply with detector voltage — going from 700 to 900 V gives roughly an order of magnitude more sensitivity, and amplifies the noise with it.
• Fluorescence is measured at 90° to the excitation beam. Anything that scatters light into the detector will look like signal.

Cuvettes

Handle by the top edge or corners. Wipe all four faces with lens tissue. Check for bubbles and for fibres, both of which scatter badly.

Sample concentration

Keep the absorbance at the excitation wavelength below about 0.1. Above that you get the inner filter effect: the front of the cuvette absorbs the excitation light before it reaches the centre where the detector is looking, so intensity stops rising with concentration and eventually falls. A sample that gives a lower reading after you concentrate it is the classic symptom.

The computer

This software dates from 2002 and runs on a correspondingly old Windows installation. Treat that machine as instrument hardware:

• Don't install anything on it, don't run Windows updates, don't connect it to the internet.
• Get data off with a USB stick, and back it up the same day.
• If the PC ever dies, the software and its licence are the hard part to replace — not the instrument. Keep an image of the drive if you can.

A · Power-on and choosing an application

1. Switch the instrument on at the front panel. The status light goes orange while it starts, then yellow while it initialises.
2. Wait for initialisation to finish before launching the software. Roughly a minute.
3. Double-click the Cary Eclipse icon on the desktop. The launcher shows the installed applications.
4. Pick the application for the job:
   • Scan — wavelength spectra, and 3-D excitation–emission maps. This is the one you'll use most.
   • Kinetics — intensity against time, with zero/first/second-order rate fitting.
   • Lifetimes — phosphorescence decay. Only available in phosphorescence mode.
   • Concentration — quantitation against a standard curve.
   • Advanced Reads — single fixed-wavelength readings.
   • Validate — instrument performance checks.
5. The application opens and runs set-up diagnostics. Let them finish.
6. Let the lamp and electronics settle for 15 minutes before quantitative work.

1.1 Setting up an emission scan

The excitation monochromator holds one wavelength while the emission monochromator sweeps a range. This is what you want when you know roughly where your fluorophore absorbs and you want to see where it emits.

Open Scan, click Setup, and work through the tabs.

Cary tab

Options tab

Reports and Auto Store tabs

1. On Reports, enter your name, and tick Parameters so the settings are recorded with the data. Tick Graph if you want the plot in the report.
2. Set peak reporting: All peaks with a sensible Threshold, or Maximum peaks for just the tallest.
3. On Auto Store, set Storage to On; Prompt at end.
4. Set Auto convert to Select for ASCII (csv). This writes a CSV alongside the native file automatically at the end of every run — the single most useful setting in this software. See section 3.
5. Click OK, then save the method if you'll reuse it.

1.2 Zero and blank

1. Put the blank — buffer or solvent alone — in the sample holder. Don't touch the polished faces.
2. Click Zero, or Commands › Zero. When it finishes, Zeroed appears in the Y display at the top left.
3. Now run a full scan of the blank across the same range. Don't skip this. It tells you where the solvent's own Raman and any impurity fluorescence sit, and it can be subtracted from your sample scan afterwards.
4. Save the blank scan with a name that identifies it as such.

1.3 Running the sample

1. Load the sample cuvette.
2. Click Start, or Commands › Start.
3. Enter a sample name when prompted and click OK. The trace draws live in the graphics area.
4. Check the signal level:
   • Over-range or a flat-topped peak — reduce PMT voltage, or narrow the slits, or dilute.
   • Noisy and weak — raise PMT voltage a step, or widen slits, or slow the scan.
   Change one thing at a time, and re-run the blank if you change anything.
5. Once the settings are right, run all samples in the set without touching them again.
6. File › Save As and name the run.

Slits versus detector voltage

Both raise the signal, but they cost you different things:

For a well-behaved bright fluorophore, keep slits narrow and let the PMT do the work. For a dim sample where you only need the peak position, widen the slits first.

1.4 Excitation scan

The mirror image: emission is fixed, excitation sweeps. Use it to find the best wavelength to excite at, or to confirm that what you're seeing is really your fluorophore — a genuine excitation spectrum should resemble the absorbance spectrum.

1. In Setup, set Scan Setup to Excitation.
2. Enter the Emission (nm) wavelength — normally the peak you found in 1.1.
3. Set Stop to the emission wavelength minus the sum of the slits. The exclusion now applies at the upper end, since you're approaching λem from below.
4. Set Start about 150–200 nm below Stop.
5. Zero on the blank, run the blank, then the sample, exactly as in 1.2 and 1.3.

1.5 Scatter artefacts — what isn't your sample

Three peaks appear in fluorescence spectra that have nothing to do with your fluorophore. Learn to place them and you'll stop chasing them.

Locating the water Raman band

Convert λex to wavenumbers, subtract 3400 cm⁻¹, convert back:

ν̃(cm⁻¹) = 10⁷ / λ(nm)

Three ways to tell an artefact from real emission:

• It's in the blank too. The definitive test, and why 1.2 matters.
• It moves when λex moves. Rayleigh, second-order and Raman all shift with excitation. A real emission peak stays put.
• It's much sharper than real fluorescence. Emission bands from molecules in solution are broad.

If second-order scatter lands on top of your emission band, either shift λex, or set the Emission filter from Open to a cut-off filter that blocks the short-wavelength order.

2.1 3-D scan — the excitation–emission map

Collects a series of emission scans, each at a different excitation wavelength, producing a surface or contour plot. Useful for characterising an unknown, or for showing that two fluorophores in a mixture are genuinely distinct.

1. In Setup on the Cary tab, tick 3-D Mode.
2. Set the excitation range — start, stop, and the excitation increment. A 5 or 10 nm increment is normal; smaller means a much longer run.
3. Set the emission start and stop as usual. Emission start must clear the highest excitation wavelength plus the slit sum, or the Rayleigh line will cut diagonally across your map.
4. Use a fast scan rate. A 3-D run is dozens of scans, and at Slow it can take hours.
5. Run the blank as a 3-D scan too, with identical settings, so the Raman diagonal can be subtracted.
6. Start the run and leave it alone.

2.2 Kinetics

Intensity against time at fixed wavelengths, with rate fitting built in.

1. Launch the Kinetics application and open Setup.
2. On the Cary tab, choose the Data mode. For phosphorescence or bio/chemiluminescence, click Options to set Delay, Gate and Decay times.
3. Enter excitation and emission wavelengths and both slits.
   • Tick Multiwavelength to follow more than one pair, then fill in the wavelength table. Useful for ratiometric probes.
4. Under Collect timing, choose Simple collect for a single stage, or Advanced collect to use different sampling rates in different stages — dense at the start, sparse later.
5. Set X Mode to Min or Sec, and set Ave. Time. Use 0.0125 s for fast reactions and longer for slow ones. That 12.5 ms figure is the instrument's floor: one data point every 12.5 ms.
6. On the Options tab, set Excitation filter to Auto, Emission filter to Open, PMT to Medium, and leave Smoothing off.
7. On the Analyze tab, choose the reaction order and the time window for fitting.
8. Zero on the blank, load the sample, and start.

Photobleaching is the thing to watch on long runs. If intensity decays smoothly regardless of what you add, run a control with no reagent — if that decays too, you're bleaching. Fix it with narrower excitation slits, a lower excitation intensity, or a longer interval between points.

2.3 Lifetimes

Phosphorescence decay only — this application is unavailable in fluorescence mode, because the Eclipse's flash lamp cannot resolve nanosecond fluorescence lifetimes.

1. Launch Lifetimes and open Setup.
2. On the Cary tab set Delay time and Gate time in milliseconds — the wait after the flash before counting starts, and the counting window.
3. Set No. of Flashes, the consecutive lamp pulses before the delay begins.
4. Enter excitation and emission wavelengths and slits, starting at 5 nm each.
5. Set Total decay time in ms, long enough to capture the decay to baseline.
6. Set No. of cycles — runs are averaged over this many repeats.
7. On the Analyze tab, choose whether to calculate rate or lifetime, tick Auto Calculate, and set the fit start and stop times.
8. Run as usual.

3 · Getting data into Excel

There are two routes. Set up the first one and you'll rarely need the second.

3.1 Automatic CSV on every run (recommended)

1. In Setup, open the Auto Store tab.
2. Set Storage to On; Prompt at end.
3. Set Auto convert to Select for ASCII (csv), or Select for ASCII (csv) with Log if you also want the parameter log written out.
4. Click OK and save the method.

From then on, every completed run writes both the native Cary file and a CSV into the current folder. Open the CSV directly in Excel — no conversion, no export step, nothing to forget at the end of a long session.

3.2 Manual export

1. File › Save As.
2. In the Save as type dropdown, choose the Spreadsheet ASCII (*.csv) option.
3. Name the file and save.

To export several traces at once, overlay them in one graph box first — they come out as adjacent column pairs in a single CSV.

3.3 Getting the file off the machine

• USB stick is the practical route. Scan the stick on your own machine afterwards; a 2002-era Windows box is not a place to be casual about removable media.
• Copy the native files as well as the CSVs. The native format keeps the full method and lets you reprocess later; CSV is one-way.
• Adopt a naming convention now — date_sample_ex-nm_em-range works well and sorts sensibly.

3.4 What's in the CSV

Two columns per trace: wavelength in nm, and intensity in arbitrary units. Multiple traces sit side by side with their sample names in the header row. Plot it as a scatter or line chart with wavelength on X.

4 · Finishing up

1. Remove the cuvette. Never leave a sample in the compartment.
2. Rinse the cuvette immediately — dried fluorophore, especially the sticky dyes, is very hard to shift later.
3. Save your data and confirm the CSVs were written.
4. Copy everything to a USB stick.
5. Close the Cary Eclipse application.
6. Switch the instrument off at the front panel.
7. Log the session in the instrument book.

Maintenance

• Keep the sample compartment clean and dry. Wipe spills at once — fluorescent residue in the compartment raises the background for everyone after you.
• Run the Validate application periodically to check performance. Do it after any service, and any time results look off.
• The xenon flash lamp is long-lived but not eternal. Falling signal at fixed settings across all samples, compared against an old reference spectrum, is the sign it's ageing.
• Only Agilent-trained service engineers should open the instrument. Always disconnect mains power before any access.

5 · Troubleshooting

6 · Quick reference

Default starting settings

Record these every run

Without them your spectra cannot be compared to anything:

• Excitation wavelength and both slit widths
• PMT detector voltage
• Scan rate, averaging time, data interval
• Solvent, concentration, temperature
• Which blank was used and whether it was subtracted