1. Using Schlenk Lines

1 Start Up

1. Make sure that all working ports on the manifold are closed, and that all joints are properly greased with high vacuum grease.
2. Attach the solvent trap(s) to the vacuum line, and seal by turning on the vacuum pump. NOTE: Each vacuum pump has its own “unique” sound when air is actively being pumped out of the system. It is important to “get to know” the regular sounds that your vacuum makes while pumping and while under full vacuum.
3. Place vacuum-sealed dewars around the solvent trap(s) and fill the dewars with liquid nitrogen (best) or dry ice/acetone slurry (only for low vapor pressure solvents — risk of vacuum pump damage) to cryogenically protect the vacuum pump.
4. Turn on the regulated inert gas flow, and adjust the flow by watching the rate at which the bubbler bubbles. (An approximate flow rate of 5 bubbles per 10 s is ideal)
5. Connect desired apparatus to a manifold port using thick rubber tubing or directly using standard taper glassware.
6. Clear the headspace (and potentially the apparatus) of residual air and moisture by performing 3 vacuum/inert gas backfill cycles: (1) open the reaction port to vacuum, wait for full vacuum, and then close the reaction port to vacuum; (2) slowly open the reaction port to inert gas, and then wait until the bubbler begins to bubble again. Rapidly opening the port to inert gas may result in air being sucked into the system through the bubbler. Close the reaction port to inert gas. (3) Repeat 3 times total.

2 Shut Down

1. Close off all manifold ports, and turn off the inert gas flow.
2. Remove the solvent trap dewar(s). CAUTION: if a blue liquid is present in the trap upon removal, it is possibly liquid oxygen (B.P. -183 °C), which can condense at liquid nitrogen temperatures (B.P. -196 °C) and can form explosive mixtures with organic compounds or cause a rapid pressurization explosion in a closed system upon warming.⁷ In the case where liquid oxygen is present, immediately replace the dewar in order to keep the trap cold, close the fume hood sash, and inform others of the situation. Leave the trap up and the system under vacuum, and leave it to slowly warm up and evaporate. Condensing liquid oxygen may be avoided by ensuring the manifolds no leaks and by limiting vacuum exposure to ambient atmosphere. A similar event can take place with liquid argon (B. P. -186 °C), and care should be taken to not pump significant volumes of argon through an LN₂ trap.
3. Turn off vacuum pump and vent the system to ambient atmosphere by opening the solvent trap vent (if applicable) or a port on the main vacuum manifold.
4. Remove the solvent traps, and dispose of any trapped volatiles into the appropriate waste container.

2. Drying Hydrocarbon Solvents/Reagents

CAUTION: Sodium metal reacts violently with water. Ketyl radicals are dangerously incompatible with some solvents, in particular halogenated solvents. Proper references should be consulted prior to choosing an appropriate desiccant for a given solvent. Follow the guidelines in Purification of Laboratory Chemicals.^8-9

1 Preparation of a “Solvent Pot” – 5 g Ph₂CO per liter of solvent.¹⁰

1. See Figure 1 for needed glassware. Under inert atmosphere, measure approximately 1 cm^5,6 Na metal (wire or chunk), cut it into smaller pieces, and place the pieces into a 500-mL round-bottom flask with a 24/40 standard taper neck joint.
2. Weigh approximately 1.25 g of Ph₂CO and place it into the 500 mL round bottom flask with the sodium.
3. Place a heavy duty stir-bar into the 500 mL round bottom flask, then seal the flask using a 180° 24/40 adapter that has been greased with a minimal amount of heavy duty high vacuum grease. Place a Keck clip over the joint to ensure a sturdy connection.
4. Remove the flask from the glovebox and evacuate on the Schlenk/high vacuum line as described in Section 1.2.6. Seal the 180° adapter and remove the flask from the line while it is under vacuum.
5. Attach a funnel to the top of the solvent pot (do not use grease!) and fill the funnel with the desired solvent. Using a long needle attached to a nitrogen line, bubble nitrogen through the solvent to partially degas it.
6. While maintaining nitrogen bubbling, slowly open the 180° adapter to introduce solvent into the solvent pot. When the solvent level in the funnel approaches the 180° adapter, close the adapter and remove the funnel.  NOTE: Solvent can be introduced in other ways, with varying degrees of air-free rigorousness. Commonly, users may choose to fill the pot while open to air and degas more later, or may choose to fill the pot with pre-dried solvent directly taken from a solvent purification system.
7. Stir the solvent pot for several hours, during which time the solution will turn deep purple indicating the formation of the sodium benzophenone ketyl radical. In the case that the pot does not turn deep purple, degas the solution (Section 2.2) and stir again.

2 Freeze-pump-thawing to degas solvent

CAUTION: Liquid nitrogen is commonly used to “freeze” solvents for freeze-pump-thaw cycles. It is highly recommended not to use liquid nitrogen unless absolutely necessary and upon consultation with supervisors. Cooling a closed system to liquid nitrogen temperatures greatly increases the chance of condensing liquid oxygen if there is a leak in the system. Furthermore, the large temperature gradient increases the chance of glass breaking due to thermal shock. In almost all cases, -78 °C is sufficient to cool a system for degassing with minimal solvent loss.

1. Start up the Schlenk or high vacuum line (Section 1.1) and attach the solvent pot to the manifold (Section 1.1.6).
2. Use dry ice/acetone cooling bath to cool down the contents of the pot to -78 °C. This may freeze the solvent depending on its freezing point.  NOTE: A common misconception of “freeze-pump-thaw” is that the solvent must be frozen to degas it. This is untrue; in fact it is easier to degas a cold liquid than a frozen solid because gas will get trapped in the solid matrix.
3. If the freezing point of the solvent is above -78 °C and the solvent is frozen, open the solvent pot to vacuum by opening the 180° adapter in order to remove the gas headspace in the pot.  Wait for the system to return to full vacuum, and then close the solvent pot to vacuum.
4. Allow the solvent pot to return to room temperature, and observe if any bubbles are forming in the solid/liquid mix. If bubbles are present, the solvent is not fully degassed and steps 2.2.2–2.2.4 should be repeated. Typically three “freeze-pump-thaw” cycles are sufficient for degassing.
5. If the freezing point of the solvent is below -78 °C and the solvent has a low vapor pressure at -78 °C, then instead of performing steps 2.2.3 and 2.2.4, the solvent pot can be opened to vacuum for approximately 5 min (or until bubbling stops) to degas the system. After 5 minutes, close the 180° adapter. A small amount of solvent may be lost to the vacuum pump trap.
6. If the freezing point of the solvent is below -78 °C and the solvent has a higher vapor pressure at -78 °C, two options are available: perform steps 2.2.3 and 2.2.4 and cool with liquid nitrogen instead of the -78 °C bath (less solvent loss, more inherent risk), or perform step 2.2.5 for shorter periods of time (more solvent loss, safer).

3. Vacuum transferring solvents/chemicals

1. Dry a 500-mL receiving Straus flask and a U-shaped vacuum transfer bridge in a drying oven (125 °C) for at least 1 h.
2. Attach the Straus flask and solvent pot to the U-shaped bridge and grease all joints lightly with grease. Attach the U-shaped bridge to the vacuum line.  NOTE: This heavy system should be supported by lab jacks and Keck clips to prevent it falling off of the line and breaking. Vacuum from the manifold may not be enough to hold it on the line.
3. Evacuate the system as described in step 1.1.6.
4. Degas the solvent pot as described in section 2.2.
5. Close off the top U-bridge valve to close off the vacuum transfer setup from dynamic vacuum. The system should now be under static vacuum with the Straus valve open and the solvent pot 180° adapter closed.
6. Use a lab jack to raise a -78 °C acetone/dry ice to cool the receiving Straus flask.  CAUTION: Cooling a closed system with liquid nitrogen is dangerous, as there is a significant chance of condensing liquid oxygen if there is a leak. While it may be faster to vacuum transfer at liquid nitrogen temperatures, it is not worth the tradeoff in safety. This practice is banned in our laboratories. If a solvent can’t be vacuum transferred at a -78 °C, it is suggested to use a better vacuum or consider thermal distillation.
7. Turn on magnetic stirring to stir the solvent pot, then slowly open the stopcock of the 180° adapter on the solvent pot, so that the liquid in the solvent pot will not rapidly boil over into the U joint.
8. Shortly, solvent should begin condensing in the receiving flask. Wait until the solvent pot is almost dried or the desired amount of solvent has been collected. Close the stopcock on the receiving Straus flask and the valve on the solvent pot, which can be refilled or used again.
9. If the solvent pot freezes during transfer, close the Straus flask valve and allow the solvent pot to warm to room temperature before continuing.
10. If the vacuum transfer is extremely slow or slows down, re-degas the system by following the protocols in section 2.2.

4. Testing the collected solvent in a nitrogen glovebox

1 Preparation of the ketyl solution

1. Weigh out 0.137 g of Ph₂CO and more 0.028 g of Na metal in a 20 mL vial.
2. Add 20 mL of THF to the vial and a small stirbar, then cap the vial and stir overnight. This should generate a deep purple solution of 0.0337 M Na/Ph₂CO ketyl radical with a small excess of Na remaining at the bottom of the vial. 

2 Titration of the collected solvent

1. Pipette approximate 4 mL of the solvent to be tested into a 4 mL vial.
2. Use a Pasteur pipette to carefully drop a single drop of purple Na/Ph₂CO THF solution in to the vial.
3. Gently stir the vial with a clean pipette tip and analyze the resulting color. Solvent that is 10 ppm H₂O or less should remain pale purple. (Not blue)
4. If the solvent will react with Ph₂CO radical (e.g. halogenated solvents such as dichloromethane), test the solvent with Karl Fischer titration or, in the case of deuterated solvents, ^1,2,3H NMR spectrometry.