A Continuous-flow Photocatalytic Reactor for the Precisely Controlled Deposition of Metallic Nanoparticles

Composites of noble metals with conductor or semiconductor materials have vast applications ranging from medical technology to energy conversion systems, but their conventional synthesis methods are not suitable for commercial production. For instance, microwave reduction is a powerful technique, but it can only process a few milliliters in each batch, while commercial production of graphene/platinum-based fuel cells would need to process several liters of suspension in minutes. To address this problem, we have developed a photocatalytic deposition reactor that is easily upscalable and can operate continuously.

In our photodeposition system, small portions of the reactants are illuminated for short, adjustable periods of time. This lets us efficiently control the nucleation and growth processes even when scaling up the reactor. To begin, cover the inner surface of a 15-centimeter-by-55-centimeter polyvinyl chloride pipe with thick, polished, adhesive-backed aluminum foil.

Space five, 55-watt UV-C lamps evenly around the inside of the pipe. Then, wrap a 0.5-centimeter-by-55-centimeter quartz tube with pieces of adhesive-backed aluminum foil to form several equally wide windows spaced evenly in the middle of the tube. Leave 2.5 centimeters of the tube exposed at each end.

Install the quartz tube in the center of the PVC tube to form the illumination chamber, and connect opaque plastic tubing to each end. Then, mount the pipe vertically in a fume hood, and fix a heavy-duty fan at the lower end for cooling, Next, set up a magnetic drive pump on a raised platform with the bottom of the pump above the top of a magnetic stir plate. Clamp a one-liter separatory funnel just higher than the pump.

Use a T-fitting and opaque plastic tubing to connect the separatory funnel outlet to the top of the quartz tube and the magnetic pump inlet. Then, insert the tubing from the bottom of the quartz tube and the pump outlet through the septum of a one-liter bottle equipped with a stir bar so that the ends of the tubing are two to three centimeters above the stir bar. This bottle will be the reservoir.

Next, connect a T-fitting to the sampling and evacuation valves and a length of opaque plastic tubing. Insert the tubing into the reservoir, and connect the sampling valve to a syringe fitting. Lastly, run a gas exhaust line from the reservoir to a water bubbler.

To begin the synthesis, add two grams of graphite and 100 milliliters of 98%by weight sulfuric acid to a 500-milliliter Erlenmeyer flask equipped with a large magnetic stir bar. Cool the mixture to about zero degrees Celsius in an ice-water bath while stirring. Then, add one gram of potassium permanganate over the course of one minute while stirring.

Let the mixture continue stirring for another four minutes. Repeat this process five more times to add a total of six grams over the course of 30 minutes. Once addition is complete, remove the bath, and continue stirring the mixture for six hours.

Then, cool the mixture in an ice-water bath while stirring for 15 minutes. Add 250 milliliters of distilled water dropwise to the stirring mixture. Remove the ice-water bath, and add 10 milliliters of hydrogen peroxide dropwise to the stirring mixture.

Then, add 20 milliliters of hydrogen peroxide all at once, and continue stirring at room temperature for 30 minutes. Centrifuge the resulting suspension at 3, 500 g for 15 minutes, and discard the supernatant. Stir the solids in one liter of distilled water for 30 minutes, centrifuge them again, remove the supernatant, and check its pH.

Wash the solids in this way until the supernatant pH reaches five. Then, combine the washed precipitate with 500 milliliters of one-molar hydrochloric acid, and stir for one hour. Wash the product until the supernatant pH reaches five.

Disperse the washed solids in one liter of distilled water, and sonicate the mixture for three hours at about room temperature. Centrifuge the resulting graphene oxide mixture for 15 to 20 minutes at 3, 500 g at room temperature three times, discarding the precipitate each time. Transfer the graphene oxide suspension to a tinted or foil-wrapped glass bottle.

Next, weigh three dry crystallization dishes on an analytical balance three times each. Shake the graphene oxide suspension well, and let it settle for one minute. Then, pour exactly 100 milliliters of the suspension into each dish.

Keep the dishes in an oven at 70 to 80 degrees Celsius until their contents are completely dry. Then, weigh each dish three times on the same analytical balance, and calculate the dry weight of graphene oxide in each dish. Determine the concentration of the suspension from the average of these values.

Then, dilute the suspension to 0.2 grams per liter, and slowly add it to an equal volume of four-molar sodium hydroxide. Reflux the mixture at 90 degrees Celsius for eight hours, and let it cool to room temperature. Wash the partially reduced graphene oxide precipitate until the supernatant reaches pH seven to eight, and then store it in a tinted or foil-wrapped bottle.

Prepare 540 milliliters of a 50-milligram-per-milliliter suspension of partially reduced graphene oxide in distilled water, and sonicate it in an ice-water bath for one hour at 40%power. Then, add 60 milliliters of analytical-grade ethanol, and sonicate the mixture in an ice-water bath for another hour. Next, add 169 microliters of 8%by weight aqueous hexachloroplatinic acid to the graphene mixture, and continue stirring at room temperature for 15 minutes.

Then, pour the mixture into the separatory funnel of the reactor and stopper it. Open the stopcock to feed the reactant into the system, and immediately close it when the funnel is empty to keep gas bubbles out of the tubing. Run the pump at 16 liters per minute, and flow nitrogen gas through the reactor at a rate that produces a steady stream of bubbles.

Stir the suspension at about 1, 000 rpm for 30 minutes to remove dissolved oxygen. Then, reduce the nitrogen flow, and turn on the UV lamps. After five minutes, increase the nitrogen flow, and connect a 20-milliliter Luer-Lok syringe to the sample line.

Lightly hold the syringe plunger to keep it from being ejected, and open the sampling valve. Collect 20 milliliters of the suspension, close the sampling valve, and detach the syringe. Open the tube evacuation valve to flush the tubing, and then close the valve and reduce the nitrogen gas flow to its previous level.

Centrifuge the sample at 10, 000 g for 10 minutes, and set aside the supernatant for metal cation concentration analysis. Wash the solids twice by stirring in 20-milliliter portions of distilled water and centrifuging under the same conditions. After that, disperse the solids in 50 milliliters of distilled water by gentle sonication for one hour.

Dissolve one milligram of ascorbic acid in the dispersion, and stir the mixture at 90 degrees Celsius for one hour. Collect and wash the precipitate as previously described, and store the washed precipitate for further characterization. X-ray photoelectron spectroscopy confirmed that platinum and gold nanoparticles were successfully deposited on reduced graphene oxide and titanium dioxide.

Deconvolution of the high-resolution platinum 4f and gold 4f peaks showed no non-metallic components, indicating that the platinum four and gold three cations had been reduced to platinum zero and gold zero, respectively. The illumination dose per exposure, or IDE, was optimized for each composite to achieve a relatively uniform distribution of small noble metal nanoparticles over the substrates. When the IDE was too high, depletion around growing particles disfavored the formation of nuclei nearby, resulting in large particles and wide size distributions.

The IDE was adjusted by altering the number of UV lamps and the exposure area of the reactor. When the IDE was too low, photodeposition was unsuccessful, owing to insufficient photoexcited electrons for stable nucleus formation, as seen here. Widening the windows to two centimeters produced the desired photodeposition behavior for the platinum/graphene composite.

Exposing the entire reactor tube increased the photodeposition rate without significantly compromising the particle size and monodispersity. The deposition process is controlled by plotting concentration time curves under various conditions to identify the optimal reactor design and IDE. Conducting the photodeposition in the linear concentration time region allows continuous collection of the product after a known circulation time.

Although we demonstrate the synthesis of platinum/graphene composites, this method and reactor can be used for deposition of other noble metals on various semiconductors too. This continuous-flow reactor can also be used for other chemical synthesis methods in which the reactions are initiated with light.