Graph of the week
Strontium project

Click on a picture to get a higher quality version

5th July 2013: We measured the temperature of the red MOT by varying the time between turning the red light off and taking an image. The expansion of the cold atom cloud as a function of time infers a temperature of 47 μK. This is an order of magnitude colder than the atoms in the blue MOT. As this is our first red MOT, we expect colder as we optimise parameters. We can, in theory, reach a temperature approaching 400 nK.

5th July 2013: Satisfied we had narrowed our 689 nm laser and stabilised it to an atomic signal, we tried for a red MOT in order to make our blue atoms colder and more dense. Here is a colour map of the red MOT.

11th June 2013: By looking at the noise on the error signal, we can infer a laser linewidth. We find that we are locked to the cavity to within 20 Hz, however this is not the linewidth of the laser as the cavity maybe jittering or drifting due to temperature changes. Typically the laser remains locked to the cavity for several days.

3rd April 2013: To narrow the linewidth of the 689 nm laser we lock our laser to a high finesse optical cavity. Here are some cavity modes we see. We lock to the first cavity mode shown, which is the TEM00 mode.

16th March 2012: Because we log the total amount of ions per run of the experiment we can extract more stastical information than by simply averaging the ion signal. The plot on the bottom left is a slice from the 2D map below. The graph above is the total occurrences of each bin count. There are 250 shots of the experiment at each spatial position. The dark blue is the total runs with zero counts, the lighter blue stripe is amount of counts with one ion, the next stripe is the total with two ions and so on. The bar graph on the right is the amount of occurrences at each count number for a particular spatial position. The data with error bar is a Poissonian distribution with the same mean.

16th March 2012: By taking several translation slices along the focussed coupling beam we can build up a 2D spatial distribution of the excited state. The state probed in this map is 56D.

13th March 2012: Using the counting and CPT we can make very accurate Stark maps, such as this one at 56D. More red colours are higher average ion counts.

13th March 2012: The saturation effect seen on 28th October could easily mask any blockade/interaction physics. Hence we want to work down in the linear part of the graph, i.e. low power. With a focussed beam and low power the amount of ions detected is very small. To make our signal-to-noise better we have gone back to counting ions. This is a spectrum of the 56D state counting the ions. The three peaks are the mJ = -2,0,+2 states, split using a quantization magnetic field.

27th February 2012: Due to the large polarisability of Rydberg atoms, the excitation volume can be controlled by applying a shaped electric field. In this case an electric field gradient has been applied to the MOT region during excitation. This shifts the wings of the spatial distribtuion out of resonance and therefore isn't excited.

28th October 2011: With the coupling laser focused much higher Rabi frequencies can be achieved This has led to a saturation type effect in the amplitude of the Rydberg signal which we are yet to explain.

21st November 2011: The coupling laser has been focused down to reduce spontaneous ionization from the regions of the MOT which don't interact with the autoionizing laser. The image on the left shows a translation signal across the narrow part of this focused laser. The image on the right shows the same data on the same axis as translation data taken along the length of the coupling laser, the aspect ratio is around thirty.

28th October 2011: Repump lasers operating on the 3P2 to 3S1 at 707 nm and 3P0 to 3S1 at 679 nm have been built. These two lasers compensate for the slight leak in the cooling transition we use for the MOT. The image on the left is of the normal MOT, the middle image is with the 707 nm light applied, and the right hand side image is with both the 707 and 679 lasers on transition.

20th October 2011: Applying a magnetic field to the cold atoms splits the degeneracy of the Rydberg transition and allows the specific mJ states to be excited. The two stretched states (&sigma+&sigma+, &sigma-&sigma-) can be easily seen.

19th October 2011: A quantization field needs to be applied so that specific mJ states can be excited to avoid Förster zero dipole-dipole interactions. This field needs to switch on as quickly as possible so that the density does not drop due to our inability to hold the atoms in a magnetic trap/optical dipole trap. The field appears to stabilise in approximately 50 &mu s.

22nd September 2011: Electron shelving in the cold atoms. The 689nm beam is locked and applied continuously to the cold atoms, this shelves the electrons in the 3P1 state, making them unavailable for probe fluorescence on the 461 transition.

19th August 2011: Applying a magnetic field to the thermal atoms used to generate the error signal of the 1S0 to 3P1 transition. This Zeeman splits the degenerate mJ states and allows us to select the magnetically insensitive state and increase the sub-Doppler slope gradient, making the lock more robust.

17th July 2011: Van der Waals interaction strength between two ground state Sr atoms (blue line) compared to the total dipole-dipole interaction strength between Sr atoms in the 50s, 60s, and 80s Rydberg state (red, purple, and green lines respectively). At interatomic separations of approximately 4 μm the interaction strength between the 80s Rydberg atoms is 19 orders of magnitude larger than between ground state atoms. In our experiment inter-Rydberg spacings of 5 - 6 μm are common. The coulomb interaction between two doubly-charged ions is shown (maroon line).

29th June 2011: Error signal of the 1S0 to 3P1 transition using fluorescence spectroscopy. The central feature is the due to the sub-Doppler Lamb dip.

1st June 2011: EIT error signals for increasing principle quantum number n states. Measured in our thermal cell. These error signals are used for locking our coupling laser in our cold atom experiments. The size of the error signal decreases with n as the transition oscillator strength decreases as n-3.

27th May 2011: Fluorescence signal of the 689 nm intercombination transition measured on a high-gain photodiode. The central peak is the 88Sr isotope, and the smaller, red detuned peak we believe is the 86Sr isotope. This atomic signal will be used in the future to stabilize the long-term drift of a high-finesse cavity, which will itself be used to narrow the linewidth of the laser.

11th May 2011: Translated integrated ion signal across our cold atoms (spatial excited state distribution) with background (spontaneous) ionization measured at each point as well. The autoionizing laser is blocked to measure the background ionization. The background ionozation is very dependent on Rydberg atom density as a plasma is formed at a threshold density vastly increasing the amount of ions.

11th May 2011: Translated integrated ion signal across our cold atoms (spatial excited state distribution) with MOT fluorescence overlayed (spatial ground state distribution). The excited state distribution is measured by physically translating a lens system which focusses an autoionizing laser beam down to 10 microns, across our cold atoms. The autoionizing beam ionizes excited state atoms with very high probability. We believe the ion signal is narrower than the MOT fluorescence because the excitation beams (probe and coupling) are narrower than the MOT.

11th April 2011: An example spectrum of the electromagnetically induced transparency (EIT) signal of the transition 5s2 1S0 to 5s56d 1D2 in our thermal vapour cell. The probe laser addressing the intermediate transition 5s2 1S0 to 5s5p 1P1 is scanned and the coupling laser is held on resonance. This spectroscopic technique is used to frequency stabilise ("lock") the Rydberg excitation laser in our cold atoms.

4th April 2011: Spectrum of the 56D state in the cold atoms with locked probe and coupling lasers - a first for us. The probe laser frequency is scanned using a double pass AOM. The feature is approximately 13 MHz wide, this is narrower than the 32 MHz of the intermediate state which indicates we are coherently transfering population into the Rydberg state.

March 2011: A laser at 689nm is shone through a hot atomic beam of strontium, and excites atoms on the narrow singlet-triplet intercombination line to the 5s5p 3P1 triplet state. The resultant fluorescence is recorded on a CCD camera. This image shows that image, and a contour plot. The laser is moving into the page, and the hot atomic beam right to left. The atoms fluoresce strongly where they interact with the laser, and the population of the triplet state is seen to decay as the atoms move right to left. From this image we can gather information about the thermal distribution of the atoms, and the lifetime of the triplet state. We can use this triplet state transition to cool strontium from milliKelvin to hundreds of nanoKelvin.

January 2011 : We've been doing some theory! By we, I mean honorary member Christohpe. This a graph of the (logarithmic) variation of the partial autoionisation width of the 5p16l to Sr^+ 5s with angular momentum,l. This was calculated numerically based on work done by Michel Poirier (Phys Rev A 38 (7), 1988). Other decay channels are to Sr^+ 4d states, which are currently not considered. Notice, the variation of the autoionization width with l is very strong, as we have observed.

December 2010 : For the long-term aim of loading strontium Rydberg atoms into an optical lattice, the atoms must be much colder than they currently are in the MOT. This requires an additional stage of cooling on the narrow (~7kHz) transition to the 5s5p 3P1 state (at 689nm), which will cool from the current 5mK to a few hundred nK. Narrowing the 689nm laser linewidth far enough is a technical challenge! This first step is a Pound-Drever-Hall (PDH) signal (blue) generated from a Fabry-Perot cavity signal (green). The signal can be used to stabilize and narrow the laser frequency.

October 2010 : We have performed sub-doppler frequency-modulation (FM) spectroscopy on the 5s2 1S0 to 5s5p 1P1 transition in our vapour cell. This provides a signal to use for laser stabilization, and the method has been optimised to yield a high signal-to-noise. The presence of the other isotopes of strontium is evident through the distortions to the spectrum to the left of the central feature. It is one of several methods we are considering to replace the polarization spectroscopy signal we currently use, to enable simultaneous stabilization of our cooling and Rydberg excitation lasers.

September 2010 : We calculate the Stark map, the splitting of atomic energy levels in an electric field, and compare it to our data (black lines). The blue and green lines are different mj states. This is still work in progress, but the agreement is good, considering we're at a complex part of the energy spectrum. This data is around the n=80 energy levels.

July 2010 : We can measure the decay of the autoionization signal at different points on the autoionization spectrum. If the sit in the wings of the 5s56d spectrum then we see the signal decay (red triangles) exponentially (red line), yielding a lifetime of 25μs. However, if we sit at the position of the central peak (see last graph), then the decay (black dots) initially decays at the same rate as in the wings (red triangles), but then there is a second component that decays with the lifetime of the 5s54f state (blue squares). An exponential fit of the 5s54f data (blue line) yields a lifetme of 64μs. A double exponential fit (black line) of the data on the peak yields an initial lifetime of 25μs, and a second lifetime of 64μs (which agrees with the 5s54f state within errors). This, along with the previous graph, tells us that population is initially prepared in the 5s56d state, but is transferred to the 5s54f state. This happens due to electron-Rydberg atom collisions due to the formation of an ultra-cold plasma.

July 2010 : If the Rydberg gas is left alone after being prepared in the 5s56d state the autoionization spectrum changes shape. It changes from the double-peaked structure shown in the last graph of the week, to a narrower single peak (black dots) (this data is taken 100μs after the Rydberg excitation). We compare this signal to the autoionization spectrum of the 5s54f state (blue triangles), and find they look similar. This data can be fitted with a simpler two-channel MQDT model (blue line) (as compared to the 5s56d state, see last graph). If we combine the six-, and two-channel models we can reproduce the data (red line).

July 2010 : This is the autoionization spectrum of the 5s56d Rydberg state. The x axis is the frequency of the laser that excites the inner valence electron, which causes the Rydberg atoms to autoionize, the y axis being this ion signal. Using a six-channel multi-channel quantum defect theory (MQDT) model we can fit the data, which is the red line.

9th June 2010 : This is a comparison of data to the simulated stark map (see graph from 20th May), for the 56D state. The red and blue lines are the simulation, the two lines for two different magnetic sublevels. The agreement is pretty good, and you can possibly even see the splitting of the magnetic sublevels.

3rd June 2010 : This graph shows the interaction energy between pairs of Rydberg atoms. The different colours are for different magnetic sublevels. The energy is relative to a pair of 56D atoms infinitely separated. The top set of curves are for the pair of 56D atoms, so at large interatomic separations one can see that this energy is zero. As they're brought together the energy increases, so the interaction is positive. The bottom set of curves is for the pairing 54F + 56P (the nearest pair of atoms in energy to our initial 56D pair). They are separated in energy from two 56D atoms by 2.5GHz (a large separation). As the spacings get smaller you can't think of the states as being pure any more.

24th May 2010 : This graph follows the purity of the 56D state as an electric field is applied. So, at zero electric field the state is purely like a n = 56, L = 2 state, but as the field increases it takes on the character of other states. We believe we see the 54F state, and interestingly by a small field of 1.5 V/cm there is 10% of this state present in the 56D state.

20th May 2010 : A calculated Stark map around the Rydberg state we've been studying (56D). For clarity I have removed the labels of the triplet states, and made the states we're interested in bold. The Stark map shows how the energy levels change in an electric field. This graph tells us that at about 4 V/cm of electric field the 56D state becomes heavily mixed with the manifold of different states.

11th May 2010 : The upper graph contains autoionizing spectra, that is the spectrum of the two-electron excitation of the Rydberg atoms. The black curve was taken straight away after the creation of the Rydberg atoms. It is broad, with two lobes, and this shape is calculable. The red curve shows the same spectrum taken 100μs later. Now we see a single peak. We attribute this to population ending up in a long-lived high angular momentum state. The lower graph shows further evidence of this. At each frequency of the autoionixing laser a lifetime measurement can be made, and the peak in the lower graph shows that at the position of the sharp peak in the spectra the lifetime is longer. The background value of about 26μs corresponds to the lifetime of the Rydberg D state.

26th April 2010 : We have studied the evolution of the autoionizing resonance over time. Straight after the Rydberg excitation there is a very broad spectrum with a complex structure. As time progresses the spectrum changes shape, and the central narrow feature becomes most prominent. We attribute this behaviour to the Rydberg atoms ending up in high angular momentum states after interaction with charged particles via cold plasma formation. This graph isn't pure data, it has been interpolated to produce a higher resolution.

15th April 2010 : We have taken spectra of the autoionizing states before using the evil pulse laser of doom. Because it's been shipped back to hell we're now using a standard diode at 408nm. We've taken a proof-of-principle spectrum (black circles) and compared it to the pulse laser data (red crosses). They are very similar, which is comforting! These experiments are interesting, because the autoionization spectra tell us a lot about the anguar momentum distribution of the rydberg atoms.

9th April 2010 : With the experiment optimised, we took lifetime measurememts of the 56D state again. This time we used autoionization as before, but also looked at the ionization from just the Rydberg excitation as well (i.e. we ionized the Rydberg atoms using autoionization, and did the same experiment allowing the Rydberg atoms to spontaneously ionize). This experiment is interesting because it tells us where the ions are coming from. At short times all of the ion signal comes from the autoionization (black squares). At longer times all of the ion signal comes from the Rydberg atoms spontaneously ionizing (blue circles). In fact, the fact that this signal rises over time is evidence of plasma formation.

19th March 2010 : We presented a large outreach project to 6th formers across the North-East, five schools in five days! The project involved measuring the change in specularity (polarisation purity) as water changes to ice, simulating remote sensing applications using radar to to measure ice cap melt. Data was logged with LabView, and this is a screenshot of the front panel.

February 2010 : We have studied the lifetime of the 56D Rydberg state using autoionization, at different Rydberg laser powers (Blue Xs: 5mW, Red +s: 10mW, Black dots: 15mW). At the higher lifetime there are two regions of decay, and the data is fit with two exponential decays. We attribute the fast decay to the true lifetime of the 56D state, and the slow decay to the fact that population is transferred to long-lived states of high angular momentum through collisions. Interestingly, at the lower laser power the decay seems significantly faster, a lifetime change of aroound 40μs to 20μs.

24th January 2010 : We extended our study of autoionization resonances in strontium to the high n Rydberg state, 56D. We were able to perform measurements on the lifetime of this atomic state, and also look at its autoionizing spectrum. In the above graph this spectrum was taken at different times after the Rydberg atoms were created. There are many interesting features to these graphs. Firstly, at later times the resonance narrows, and the centre value moves towards the "bare ion line", which is the excitation resonance of the strontium ion. Also the signal in the wings, where the pulse laser isn't resonant, changes as the Rydberg atoms begin to spontaneouly ionize.

18th January 2010 : We have been performing "autoionization spectroscopy". After we have excited strontium to a Rydberg state we use a pulse dye laser to excite the second electron. This ionizes the atom extremely efficiently, we can ionize all of our sample. This technique is a very precise measure of Rydberg state population. Because performing autoionization on cold atoms is a new technique we have taken some spectra. The above graph shows the spectra of the 19D and 20S Rydberg states as the pulse laser is scanned in frequency. These features are extremely broad, spectral features tend to be tens of MHz, whereas these are on the order of THz. The D state is much broader, and has more structure.

15th December 2009 : We now have accurate data on two widely separated regions of the Rydberg spectrum. The graph shows the amount of ions present at different times after the creation of the Rydberg atoms. The 4μs excitation pulse of light that creates the Rydberg atoms lasts from 0-4μs. The red curve is at a principle quantum number of n=20, and the black is at n = 80. This graph is a good indicator of the difference between these different states. At n = 20 (red) the coupling is much stronger, so the signal is higher. At n = 80 the signal is lower, but persists for longer as the Rydberg atoms mix into a plethora of other states, some of high angular momentum which have long lifetimes.

11th November 2009 : First our cold atoms are excited from the 5s5p 1P1 state to the 5s20s 1S0 Rydberg state using our purple laser at 420nm. As we have investigated thoroughly before this produces ions, probably through the Rydberg atoms colliding. The ammount of ions you get as you scan the 420nm laser across the resonance is shown in the red trace. However, of a purple laser at 408nm is also present you see a massive increase in ion signal (black trace). NOTE: the red trace has been magnified by 20 times, so is much smaller than the black trace. This happens because the 408nm laser excites the second electron in the strontium atoms. This situation is unstable and the atoms ionise directly.

23rd October 2009 : Data has been taken for the singlet S and D series' at principal quantum numbers of n = 18 and n = 20, and compared. The data shown above is a measure of how many ions we detect at certain times after the excitation of Ryberg atoms. The Rydberg excitation lasts from t=0-4μs. The behaviour is similar for the 18D and 20D states, however it is markedly different for the 18S and 20S states. One clue could be that the 18S transition is close to an autoionising resonance, where the laser can also excite the second electron, ionizing the atom.

13th October 2009 : By solving the three-level optical Bloch equations for our system (5s2 1S0 → 5s5p 1P1 → 5s18d 1D2) we have found the time evolution of the atomic population in each state. We find that actually quite a large fraction of our atoms end up in the Rydberg state, about 42%. This means that the density of the Rydberg atoms is of the same order as the density of our initial atomic cloud (about 1010 atoms per cubic cm). We have measured that we only lose a small fraction of our atoms through the Rydberg excitation (a few percent), so this means that very few of our Rydberg atoms are ionizing.

30th September 2009 : In studying the Rydberg state 5s18s 1D2 state we decided to study the effect of atomic density on ion signal. This graph shows the ion spectrum at a density of 4.4x1010 atoms per cubic cm (black) and 2.7x109 atoms per cubic cm (red). Interestingly the ion spectrum at a lower density it considerably narrower. We don't know why yet!

18th September 2009 : We are now using a laser at 420nm to access lower lying Rydberg states. We perform a two-photon excitation via the 5s5p 1P1 to access the 5s18d 1D2 state. In our previous work on EIT in a thermal beam we believe we may have hit an autoionizing resonance, where once the Rydberg excitation has been made the core electron is subsequently excited. This causes the atom to ionize. When we perform this excitation in the cold atoms we see a large drop in trapped population (see last graph) and also a large ion signal. This graph shows the time evolution of this ion signal. Ions are rapidly created (in a few μs), and then the signal dies away in some tens of microseconds. Compare this to previous graphs looking at very high Rydberg states, where the ion signal persists much longer than the 100μs timescale.

28th August 2009 : When we expose atoms in out magneto-optical trap to light resonant with the 5s2 1P1 → 5s18d 1D2 Rydberg state transition (at 420nm) we see a massive loss of population from the trap. In the above graph this is shown by the huge drop in fluorescence as the 420nm laser is scanned across the resonance. It is so marked that you can see with your eyes as the 420nm laser blasts through the trap, shown by the CCD camera images on the right (above: no 420nm light, below: with 420nm light). This could happen simply because we are exciting a very large fraction of the atoms to the Rydberg state, or it could be because we have hit an "autoionizing resonance", where the 420nm light excites the second strontium valence electron and the atom ionizes.

14th August 2009 : We used our Stark shutter technique and a 10μs pulse of on resonant "probe" light to produce Rydberg atoms at a set time. We then scanned an electric field pulse across this excitation to detect ions created from the Rydberg atoms. This gives us an indication of what is happening as the Rydberg atoms ionise. The different graphs are for different amounts of power in the probe beam, you get more signal when there's more power, because more atoms are excited to a state that can then be excited to the Rydberg state by the 412nm laser. The circles are the actual data, and the dotted lines are numerical fits. In all cases there is a fast linear rise, indicating the speed at which the Rydberg atoms ionise (something like 5μs), and then an exponential drop-off in signal, indicating how long the ions hang around for. We don't know why this doesn't go to zero yet!

10th August 2009 : In previous experiments the laser at 412nm has been on the whole time, so while the atoms are trapped in the MOT (most of the time, periods of about 30ms) Rydberg atoms are being created, leading to a build up of charge. Ideally we would turn off the 412nm laser until we need it, which would enable us to turn the MOT off, apply a pulse of resonant "probe" light at 461nm and simultaneouly a pulse of 412nm light. This would allow us to know exactly when Rydberg atoms are made.

We unfortunately do not have equipment available to do this, so came up with a different technique which we've dubbed "Stark shuttering". While the MOT is on we apply a uniform electric field to the atoms. This shifts them away from resonance with the 412nm light, so no Rydberg atoms are made while the field is on. The black curve above shows a (probably saturating) ion signal, taken by applying a pulse of electric field at some time after turning off the MOT. We can see a peak corresponding to accelerated charge. However, in the red trace an electric field has been on while the MOT was on, which we turn off with the MOT. Now we see no ion signal, indicating that no ions (and hence Rydbergs) have been created.

5th August 2009 : This is a Stark map built around the n=80 Rydberg states of strontium. The black curve at the front is a spectrum taken in (roughly) zero electric field, showing several peaks corresponding to transitions to different angular momentum states. As an electric field is applied (going backwards in the plot) these peaks move, split, broaden etc. in interesting ways that, in theory, one can predict. The alternating colours are only for clarity.

7th August 2009 : The Rydberg spectrum was extended and filled in between n=56 to n=125, locating many more lines than previously. At low n (corresponding to deeper energies below the ionisation threshold, or to the right of the graph) the different series are clearly defined. As you head to higher n, closer to the ionisation limit, the lines mix and the series' are no longer discernable. For an explanation of this type of plot see graph of the week from 9th June 2009.

28th July 2009 : The black curve shows the level of fluorescence from the MOT, and is proportional to the number or atoms in our trap. The red curve shows the ion signal. The trap and Rydberg laser are on continuously. At t = 0 an electric field that is roughly 3.6V per cm is applied, which points towards our detector, and lasts 50ms. When it is first applied there is a large spike of ions detected, before the background level of ions decreases (we have emptied them all). At the same time the trap fills up to a higher level. We think this happens because the electric field shifts the Rydberg atoms away from resonance with our laser, and hence the Rydberg fraction of our atoms is no longer being lost from the trap.

21st July 2009 : Our MOT is turned off, and a pulse of electric field is applied, creating an ion signal. The variation in the size of this signal with the time it's applied is shown in the black curve. 30μs after this electric field pulse a second field pulse is applied, and the ion signal is shown in the red curve. The size of the first peak is indicitive of the number of ions around, and it should clear out most of the ions. The size of the second peak is indicitive of how much the reservoir of ions has refilled. We believe the ions come from Rydberg atoms, so the fact that the red curve decays much faster than the black shows that their lifetime is short compared to the time that the ions hang around.

15th July 2009 : After our MOT light has been turned off two electric field pulses are applied, separated in time, to direct charge towards our detector. This graph shows the ratio of the height of the second peak to the first, and its variation in time. Since the first pulse is big enough to empty all the charge, the fact that the signal recovers and stays around means that charge is being created, and hangs around for a long time.

8th July 2009 : This graph shows a saturated absorption spectrum of strontium, on the 5s2 1S0 to 5s5p1P1 transition, taken in our strontium vapour cell. The computer was used to control the frequency of our laser (at 461nm), and we read the wavelength of the light out on a wavemeter. We know what shape this spectrum should be, so we can check that the laser control and wavelength measurement are correct. The black dots are data, the red line a theoretical fit, and they're in good agreement.

2nd July 2009 : This graph is taken by the same method as the previous one, except that the taking was fully computer automated. The 412nm laser is stepped across the Rydberg resonance, producing ions as seen in the red trace. Using the change in the flourescence signal from the atoms the number of atoms can be calculated at each frequency, as shown in the black trace. This graph suggests we are exciting some 100,000 Rydberg atoms.

25th June 2009 : The red line shows the ion signal as the purple (412nm) laser is stepped in frequency across a Rydberg resonance (a 1D2 line at about n=70). At the same time there is a series of excitations of our trapped atoms to the intermediate state (from which they can be excited to the Rydberg state), this is shown by the sharp lines in the black trace, which is a measure of the fluorescence from the atoms. The fact that this black signal dips when there is an ion signal means that there are less atoms to fluoresce, so we are making enough Rydbergs to deplete the trap! (The Rydberg atoms cannot be held in the trap, so are lost).

9th June 2009 : If you pick out the lines in the spectrum (i.e. the previous graph) and calculate the effective principle quantum number n*, then by plotting the value modulo 1 you essentially have a measure of the quantum defect δ = n - n*. At high n the different series tend towards a constant and unique value of δ. You can see five clear series, so we can begin to identify which resonance is which in our Rydberg series.

29th May 2009 : Taken in the same way as the previous graph, this is several laser scans stitched together. These lines are somewhere around n = 70-72. You can really start to see patterns emerging now, as you would expect. There are two series that can be directly excited, the 5sns 1S0 and 5snd 1D2 series. However, due to mixing at these levels triplet series can also be excited, as can transitions to P and F states.

29th May 2009 : Atoms held in a MOT were excited into the Rydberg region of energy levels via a two photon excitation (the first photon from the MOT beams, the second from a purple laser at 413nm). The 413nm laser is slowly scanned over several gigahertz of frequency, and an electric field is pulsed on many times during this scan. The electric field accelerates the charge towards the detector. It is hard to say which resonance is which, due to the complexity of the energy levels in high n states. However, we are in the region of n = 70-80.

13th May 2009 : This graph is essentially the same as the previous graph, however instead of counting the number of ions that are registered by the electronics, the signal direct from the detector is measured and averaged to build up an integrated signal. This method means you are not limited by the speed of the detector, however you lose gain. As the grid voltage is increased you see the averaged signal increase, with no saturation problem.

13th May 2009 : These peaks show the number of counts (ions) registered in each 1μs bin in an experiment where ions were made and then accelerated towards our detector using a pulsed electric field. In front of the detector is a mesh ("grid") that can be charged to help direct the ions. As this grid voltage is increased you can clearly see the signal saturate, which is not observed when you count "by eye" on an oscilloscope, showing that our detector is flawed. Though we are getting much less than one event per pin on average, the events are arriving so close together (sub-μs) that they are not all being counted.

7th May 2009 : These pictures show the electric field surfaces around the region where our trapped atoms are. The cloud of cold strontium is dead in the centre of these plots, inhabiting a region about 1mm square. Our detector lies in the negative x direction. What these pictures tell us is that the ions can be accelerated towads the detector, without the field pushing them in other directions, and also without a large gradient.

30th April 2009 : This experiment involved pulsing a beam of light resonant with the strontium atoms in our trap for 30μs (leading to ionisation), followed by a pulse of electric field towards our detector of a duration shown on the x-axis. The y-axis is the peak average number of ions detected every μs. For the first 25μs the field is ramping to its maximum value, and from here onwards it's at about 7Vcm-1. We interpret this graph as saying there is an optimum field pulse length, after which some constant fraction of the ions get "sucked" onto our electric field plates.

Content © James Millen, Durham University 2009