This post provides a journey through the entire experiment, including sample preparation, experimental design and data evaluation.

Coral sample preparation: 200 micron thick embedded coral slices

1) Sample labeling
→ every sample was assigned with numbers: 500 - 505
→ every sample side: a, b, c, etc.
→ p = Pocilopora
→ s = Stylophora

1a) Samples measured (highlighted in yellow) with XRD, XRF, absorption –> RADIATION DAMAGE (KATREIN):

1b) Samples measured with XRD, XRF, absorption –> FLOW / NO FLOW (ISABELA):

2) Raw sample cutting
→ Untreated corals were cut into 1-1.5 mm thick slices:
1. Embed end of sample in thin layer of pmma, which then can be stuck to the transparent rectangle that fits in the suction plate of the bandsaw, to better control quality and thickness of slices of raw coral
→ Bandsaw (Name, company,location):
EXAKT dia-plus Walter Messner GmbH, Barsüttlerweg 6, 22113 Oststeinbeck, Germany

3) Embedding
Each sample was potted in methyl-methacrylate (Technovit 4071, Kulzer GmbH, Germany). Firmly secured at the potted ends, samples were then sectioned cross-sectionally into ca. mm-thick sections. Each section was embedded individually (Technovit 4071, Kulzer GmbH, Germany), so as to infiltrate otherwise closed internal pores.

→ Embed these slices in pmma, filling the mold to the top.
→ Side A = bottom of embedding mold = side of puck with sample near surface
→ Side B = top of embedding mold = far from sample
→ End up with a puck, where thin slice of sample is at one end, and the rest is pmma.

4a) Sample polishing short for paper
Embedded pucks containing the sample section were then ground to expose the coral skeleton material and polished (EXAKT 400 CS, EXAKT Advanced Technologies GmbH, Germany) on both sides, to produce 200-400 µm-thick highly polished plane-parallel slices. Specifically, blocks were first roughly ground down to ca. 500 um thick (800 grit), making sure that the coral skeleton was exposed on both sides of the puck. Then, each side was polished using the following protocol: 1. 1 min at 1000 grit with water 2. 1 min at 1200 grit with water 3. 3 minutes at 2500 grit with water 4. 3 minutes at 4000 grit with Na2CO3 All polishing steps were done with xx g weight. The final step at 4000 grit was done with Na2CO3 to prevent surface ACC dissolution (YUT Gong et al., PNAS 2012). Samples were then wiped (Kimtech wipe with Na2CO3), dried (pressurized air), and stored in a clean membrane box.

4b) Sample polishing in detail: Polish side A
→ Can work in batches of ~6 samples.
→ Polish side A (removes 0.1 mm):
1. Manually sand side B with 800 grit paper, just to make sure surface is clean and smooth and will stick well with tape. Wipe dust off with wet tissue and dry surface with compressed air duster.
2. Stick samples (side B stuck, side A exposed) onto transparent rectangle
3. Polishing standard procedure (Removes 0.1mm total):
- 1000 grit, 1 small weight, ~2 minutes (or more if needed) to expose surface
- 1 min @ 1000 grit, 1x small weight
- 1 min @ 1200 grit, 1x small weight
- 3 min @ 2500 grit, 1x small weight
- 3 min @ 4000 grit, 1x small weight

4c) Polish side B
→removes 0.1 mm, Plastic rectangle thickness + 1 layer of tape = 2.2 mm → target thickness = 2.2 + 0.3 = 2.5mm
→ Use bandsaw to cut off excess pmma from side B, leaving pucks of ~1.5mm thick (with the coral slice inside).
→ Grind side B (800 grit, 2x small weitht + large weight) until sample is 0.3 mm thick (0.3 + 2.2 = 2.5 sample plastic tape unit). In polisher machine, can use the meter knob to set thickness at which polishing will stop. Make sure to zero without sample.
→ Stick samples (side A stuck, side B exposed) onto transparent rectangle: To allow eventual detachment of sample without damaging delicate slice, stick it with thin slivers of tape only on two sides of the puck where there is just pmma.
→ This will be slow, goes faster if use little water. Can start with 320 grit and move to 800 once approaching target thickness.
→ Then do standard polishing as above
→ Remove samples from plastic rectangle. Use fine dentistry tools to slowly lift tape off of plastic rectangle.
→ Wipe slice with kimwipe and solution of filtered 22g/L Na2CO3 (in fridge in front of the door of the room where Keyence is, in glass bottle with blue cap, labelled…you will see it).

Note: Not further polished with diamond/silica/alumina paste because in test sample this standard polishing was enough to reach < 1um roughness on side B (last polished surface).

In the end, the polishing machine is not that precise that I cut the samples with the band saw as thin as possible. If I chose to cut the sample too thin, the band saw makes funny things and the sample just breaks. After cutting I polished with 800 grit polishing paper and polishing machine till a thickness of approximately < 1 mmm (determined with blank eyes). Then increasing stepwise the grit of the polishing paper (see polish side A) till I am happy with the surface including thickness. At the end I measured the thickness precisely and noted the size in the table above.

5) Before beamtime measurements:

Samples were measured with light microscopy

6) At the beamtime:

Participants: Katrein Sauer, Paul Zaslansky, Isabela Vitienes, Shreya Ray

Samples were measured within a diaphrame samdwiched within two yellow Kapton films / polyimide films (Kapton-Folie / Polyimid-Folie, Imid-group: -CO-NH-CO- –> electric isolator, thermal stable and chemical- and radiation resistant).

Energy: 17 keV
beamsize: 50 microns\

Plan:

1. Mapping the entire samples with a stepsize of 100 x 100 µm
2. choose two ROIs with a stepsize of 50 x 50 points (=highest possible resolution defined by beamsize) with 1s or 2s exposure time
3. Irradiate several choosen spots over and over again, and collect XRD patterns (Debye rings) over time

Close the Hutch:

• press green button on the wall
• close door
• rotate key right beside the door and take it with your
• put key into beamline lights which shows green light if everything works well (MySpot (NOT BAMline!)) and rotate key 1/4 to the right
• press leverage (Hebel) up and release. It will beep and every diode in that line should light up green now

Open the Hutch:

• rotate key from 0–>1 and take it with you to the pace right beside the door
• enter key and bring it to position 1
• press green button
• door will make a noise, you can open it

**Helpful commands:**    -   beamstop_out    - PAUSE/STOP: \
P [=pause] \
r [=resume]\
Strg + C [=der totale Abbruch]
- curs [= to go anywhere within ascan with motor] \
 m [= move motor to that position]
 q [= quit]
- wm kox [= where motor (position) kox]
- wm mz [= where motor (postition) mz]

7) Additional measurements after beamtime:

1. Sample s504d for XRF to look into Magnesium (PTB beamline (Adrian))
2. Leica light microscopy
3. SEM-BEI
4. Laminography (Isabela)

8) Evaluation

Transmission / Absorption (diode1):

• diode1 = absolute absorption, good calibration, gives photons / s
• cyber = transmission (not as good calibrated as beam + has to go to through air and through the sample –> absortion of both. Therefore: transmission = air / sample)

X-ray Absorption Images

Transmission images from Spec-files (BESSY, found wihtin one of the last columns)

Transmission (T) transformed into Absorption (A):
A = 1-T
–> T was determined with Fiji, looking for the highest intensity value within histogram and this is defined then as “100% transmission = 1 in formula”:
a) 100% = 593751.44
As -A = T - 1 and T = our image:
with Fiji:

→ click on transmission image → process → math → substract 593751.44 → afterwards multiply with (-1)

XRF FLUORESCENCE:

Measured elements as part of coral samples:

Measured elements as part of background / beamline setup / contamination:

The arsen (K_alpha) and lead (L-line) energy are both included in one peack as this peak is broad enough to carry both. Therefore, it is not clear which element it really is exactly but most likely LEAD!

From measurement set68 till set74: Here is a mistake in naming: Instead of sample p501c, it actually is the sample s503a and must therefore be set68_s503a,…,set74_s503a, and also set76_s503a,…,set89_s503a!!! The same is true for XRD!

XRD X-Ray diffraction (WAXS):

Energy: 17 keV
beamsize: 50 microns
SSD:
Detector: Eiger 9M
Diaphrame size:
mz = 0 - 25 (window inner size = 24 mm)
koi = -20 -0 - +25 (window inner size = 36 mm)
–> Richtung -20: fährt man zur Wand

CALIBRATION OF s504d_Corond_1
Energy: 16.9506007keV
Lamda: 0.731444 Angstroem
Pixel size x, y: 75 x 75 microns²
SDD: 35.2452 cm
Center X: 1619.92 pixels
Center Y: 1695.28 pixels
tilt a: -0.39506460 degrees
tilt b: -13.292978 degrees

ALIBRATION OF s504d_Corond_2
Energy: 16.951478keV
Lamda: 0.731406 Angstroem
Pixel size x, y: 75 x 75 microns²
SDD: 35.1977 cm
Center X: 1619.87 pixels
Center Y: 1695.34 pixels
tilt a: -0.39132528 degrees
tilt b: -13.171041 degrees

AVERAGE s504d Corund 1 and 2
Energy: 16.95103935
SSD: 35.22145
lamda: 0.731425
Center x: 1619.9
Center y: 1695.3
tilt a: -0.393
tilt b: -13.25\

Mask off:
Valid pixels
< 2000
apply mask

Calibrated SDD for s504d:
set03: kox = 3.5, SDD = 35.21159
set04: kox = -3.5, SDD = 35.2278097
set05: kox = 3.2, SDD = 35.212053
set06: kox = -2.2, SSD = 35.225029260768295
set07: kox = 5.4, SDD = 35.20718782223171
set08: kox = -3.6, SDD = 35.2280414517683
set09: kox = -3.7, SDD = 35.228273158768296
set10: kox = -3.5, SDD = 35.227809744768294
set11: kox = 4.3, SDD = 35.20973659923171
set12: kox = 3.5, SDD = 35.21159025523171
set13: kox = 2.2, SDD = 35.21460244623171
set14: kox = 5.3, SDD = 35.20718782223171
set15: kox = -4.4, SDD = 35.229895061426895
set16: kox = -4.3,SDD = 35.229663400768295
set17: kox = -4.5, SDD = 35.2301268147683
set18: kox = -4.1, SDD = 35.22919998676829
set19: kox = -3.8, SDD = 35.22873652642689
set20: kox = -4.4, SDD = 35.229895061426895
set21: kox = 2.6, SDD = 35.21367557189031
set22: = 2D mesh
set23: kox = -0.3,SDD = 35.22039512097683
set24: kox = 0.8, SDD = 35.21784634404634
set25: kox = 0.3,SDD = 35.219004879023174
set26: kox = -3.5,SDD = 35.2280414517683
set27: kox = -3.4, SDD = 35.2275779914269
set28: kox = -4.0, SDD = 35.22919998676829
set29: kox = -4.4, SDD = 35.229895061426895
set30: kox = -4.6, SDD = 35.230358521768295

All .h5-files were transformed to .tiff for further evaluation using Fiji.

From measurement set68 till set74: Here is a mistake in naming: Instead of sample p501c, it actually is the sample s503a and must therefore be set68_s503a,…,set74_s503a and also set76_s503a,…,set89_s503a!!! The same is true for XRF! No damage visible in aragonite, i.e., no calcite was created.

### Looking into single damaging points

set03: SUM,MED, and MAX of 120 images at the beginning compared to 120 images at the end doesnt give any phase changes, nor creating a phase such as Calcite. Only when zooming in, there is a tiny peak where calcite is supposed to be expected. However, peak intensity does not change at all in terms of creating calcite!!!

However, it might be the case that the intensity of the aragonite increases, as a result of radiation damage? Could it be possible, that the calcite is eradicaed as a result of radiation damage? NO: SO FAR: The Intensity of aragonite as well as calcite increases over time, and then slightly seems to deminish and decreases.

## set02: Mapping calcite-channels
(104):
center: 1924
left: 1910
right: 1940

aragonite channels

**9) XRD Integration from Ivo**

Ivo only took one corundum for all the samples: NEEDS TO TAKE ALL CORUNDUMS LEFT AND RIGHT OF THE Samples

INTEGRATED “CALCITE” IMAGES ACTUALLY BELONG TO AN ARAGONITE PEAK!!!

We see already huge and very few single calcite peaks (~ 1µm-range?!)