SrCs4(PO3)6 synthesis

[Brian Joy]

With a ton of help from fellow forum member crystalgrower, I've just completed my first SrCs₄(PO₃)₆ synthesis run.  I haven't had a chance to look at the crystals under the microscope yet, but the habit appears to be consistent with the crystal system (hexagonal).  The run produced several thousand small, acicular crystals that I'll intend to use as seed crystals in subsequent runs.  In the photo below, the width of the field of view is about 35 mm.  The largest crystals are perhaps 4 mm in length but very narrow.  I'll probably start the next run to produce larger crystals late tomorrow.

[Brian Joy]

I haven't had a chance to look at the crystals under the microscope yet, but the habit appears to be consistent with the crystal system (hexagonal).

So apparently I posted too quickly.  I had a sneaking suspicion that the symmetry looked too low, and I just verified this under the microscope.  The crystals appear to be monoclinic and may be Sr₂Cs(PO₃)₅.  This could also explain the relatively low yield.  I'll probably start a second attempt later today.

[Brian Joy]

I’ve attached a couple of BSE images of the Sr-Cs phosphate that I produced in my first run.

The compound has symmetry no higher than monoclinic and appears to be the only phase present aside from a few grains of a hydrous Cs-Al phosphate (due to use of an alumina crucible).  It possesses at least two perfect cleavages, and these cause some problems when polishing randomly oriented grains.  Both simple and polysynthetic twins appear to be present.  The compound has been boiled, ultrasonicated, ground, and polished in H2O and shows no evidence of solubility in it.

I’ve analyzed it quantitatively, but the combination of a lack of a reliable Cs standard and significant atomic number and/or absorption corrections for Sr and P make me reluctant to assign a formula unit at this point.  What I’m willing to state is that the formula unit is likely bounded by Sr₂Cs₃(PO₃)₇ and Sr₃Cs₅(PO₃)₁₁.  The first formula corresponds to wt% SrO = 18.39, wt% Cs₂O = 37.52, and wt% P₂O₅ = 44.09, while the second, wt% SrO = 17.31, wt% Cs₂O = 39.23, and wt% P₂O₅ = 43.47.

The compound is beam sensitive and shows clear evidence of Cs migration under a focused beam, even at 2-3 nA beam current.  When I performed the analyses, I used a 15 kV potential, 5 nA current, defocused the beam to 10 microns, and counted for 10 s peak and a total of 10 s background.  I intend to analyze it again at different beam currents to see if the results differ significantly, but I won’t be able to do this until late in the week.

The initial thought was that the topology of the phase diagram for the Sr(PO₃)₂-CsPO₃ system should mimic that of the Ba(PO₃)₂-CsPO₃ system.  In the latter system, the two intermediate compounds are BaCs₄(PO₃)₆ and Ba₂Cs(PO₃)₅.  So either I’ve produced something metastable, or the topology of the Sr(PO₃)₂-CsPO₃ system departs from that of the Ba-bearing system.  I have a second run in the furnace right now.  I kept the first one in the furnace for 22 hours, and I’ll let this one run for perhaps ~100 hours just to see if the results are time-dependent.  If I get the same results, I’ll try to grow larger crystals in run 3.  I’ll likely try a run in the Ba(PO₃)₂-CsPO₃ system for comparison.

[Probeman]

I ran Probe for EPMA in simulation mode (using the built in Penepma code) for a quick look at any possible spectral interferences.  Looking at the EDS spectrum, the Cs L family looks fine all by itself, but P and Sr might interfere with each other as seen here:



So I then ran a simulated WDS scan and here are P and Sr looking very well separated:





Obviously Cs would be the main interest, but it is good to see that it could be a useful standard for the other elements as well.  That is, as long as it's stable under the beam!   

[Brian Joy]

I removed my second run from the furnace this afternoon after 90 hours.  The results appear to be entirely consistent with what I got in the first run, and so it appears likely that the topology of the Sr(PO₃)₂-CsPO₃ system is in fact different from that of the Ba(PO₃)₂-CsPO₃ system.  I don't think it's really worth pursuing production of a beam-sensitive compound of uncertain stoichiometry.  Andrew Locock has suggested CsCoPO4 as a possibility, and perhaps this route would be more productive.

[crystalgrower]

Hi Brian,

In addition to my further suggestions offline, there is one observation.  It took me 5 tries to get this mixed business, and then I hand picked the desired phase from the mix to seed subsequent batches.

There is one other factor that is not within your control:  Nucleation is controlled by crucible material   If a Pt crucible is not available, then a run with a piece of Pt wire should be tried if at all possible.

[crystalgrower]

Hi Brian,

This is where a departure from the published method may be useful.  It looks like you have gross supersaturation which is not likely caused by anything you did.  The crystal structure of BaCs4(PO3)6 was published in Acta Crystallographica, use your library search engine.  A Durif was one author.

At this point it would make sense to increase the amount of 85% H3PO4 that you start with.  Depending on the headspace in your crucible, at least 2X, or 3X if possible.  The extra H3PO4 acts as a flux, it does not affect the chemistry.  Heat the crucible and acid for 2 hours at 150C to dry off all water.  Let it cool to 100C to add SrO.   Then stir in Cs2CO3 very slowly and cautiously to allow CO2 to come out. 

Top up the H3PO4 to half full after all Cs is in. Give it an hour at 150C and then the slow gradient to 450C.

You do not need to increase cooking temp above 450C nor the 24 hour cooking time.  They are not the problem.  Good luck.

[Brian Joy]

Hi Brian,

This is where a departure from the published method may be useful.  It looks like you have gross supersaturation which is not likely caused by anything you did.  The crystal structure of BaCs4(PO3)6 was published in Acta Crystallographica, use your library search engine.  A Durif was one author.

At this point it would make sense to increase the amount of 85% H3PO4 that you start with.  Depending on the headspace in your crucible, at least 2X, or 3X if possible.  The extra H3PO4 acts as a flux, it does not affect the chemistry.  Heat the crucible and acid for 2 hours at 150C to dry off all water.  Let it cool to 100C to add SrO.   Then stir in Cs2CO3 very slowly and cautiously to allow CO2 to come out.  Give it an hour at 150C and then proceed with a slow gradient to 450C.

You do not need to increase cooking temp above 450C nor the 24 hour cooking time.  They are not the problem.  Good luck.

Hi Irene,

I'll give it another try this week, probably Wednesday.  I tried growing BaCs₄(PO₃)₆ using the same amounts of starting materials as Masse and Averbuch-Pouchot (and with Pt wire in the crucible), but I failed to produce any crystals at all.  My crucible is plenty big enough to hold quite a bit more H₃PO₄, so I'll add extra in this next run.

[crystalgrower]

Brian, my suspicion is that some early experiments had something present to trigger nucleation.  Some materials appear to inhibit nucleation.  Crap shoot either way.

Cleaning the crucible with H3PO4 that you then discard takes care of this.  It is the only reliable way that I found.