Adsorption of Benzene using Activated Carbon - Increase in concentration?

[FaNgS]

Adsorption of Benzene using Activated Carbon - Increase in concentration!?

I'm not quite sure where to post but here it goes.

I'm performing an adsorption experiment of benzene from solution using commercial activated carbon. It is expected that the concentration of benzene would decrease with time since it gets adsorbed on the carbon, but in my case the concentration increased!

I'll list the steps I used for my experiment first and all the details for preparation, experimentation and analysis:

I'm performing an adsorption experiment of benzene from solution using commercial activated carbon. And I'm using a gas chromatography fitted with a flame ionization detector for the analysis and an auto-sampler. Also I'm using the internal standard method for making a calibration curve and determining my solution concentrations after adsorption.

I've prepared a stock solution of benzene with ethanol as my solvent. I then made three standard solutions of different known concentrations by diluting the stock solution. I'm using glass volumetric flasks with glass stoppers.

I used ethyl acetate as my internal standard which is also dissolved in ethanol. Similar to the benzene solutions, I made a stock solution and diluted it down to make a single standard solution of known concentration. The internal standard is not introduced into the adsorption experiment, it is only added for analysis.

For the experiment part, I use 40mL glass vials with a PTFE screw caps and rubber seals. In the glass vial I added 50mg of activated carbon and filled it up with my standard solution and ensured minimum headspace to avoid any vapor losses. After specific times, I filter this solution and centrifuge it to discard any activated carbon particles.

To create the calibration curve, I add 0.5mL of my internal standard and 0.5mL of my benzene standard solution into a 2mL glass vial with PTFE screw cap and rubber septum. This is then injected into the GC and I record the area of the ethyl acetate and benzene peaks.

Taking the ratio of the area of benzene to the area of ethyl acetate as my y-axis and the concentration of my benzene standard solution on the x-axis I use a spreadsheet to calculate to the equation of the line. Using this equation I'm able to calculate the concentration of my experimental samples after adsorption.

This is where things get weird. I did an experiment using 400ppm benzene solution and after 24 hours, I checked the concentration expecting the concentration to decrease or remain the same (logically it should decrease). BUT it increased!

Can anyone help explain this? I've repeated the experiment several times and each time the concentration increases to higher than the original solution concentration.

 

[JohnRC]

A few possibilities suggest themselves:
1) activated charcoal as commercially presented might have some aromatic type impurities. Did you run a control blank experiment to see whether benzene was present when you simply exposed your (fresh as presented) adsorbent to the ethanol solvent you were using?
Did you try running the adsorption experiment in activated charcoal that had been pre-leached with the ethanol solvent you were using?

2) Ethanol is hardly the ideal solvent to use in an experiment like this, since ethanol is also adsorbed to some extent onto the active sites of a charcoal adsorbent. Could your results be explained if the adsorption of ethanol were actually stronger than that of benzene? Could you run your experiment with water as solvent? I know that benzene is only sparingly soluble in water, but it is more soluble than a typical aliphatic hydrocarbon, and low solubility is generally a good thing in adsorption studies. Note that you do not have to use water as the GLC carrier; the chromatography step of your experiment is essentially distinct from the adsorption step, and you could do the chromatography by starting with pure water and ramping up an admixture of ethanol.

 

[FaNgS]

Thanks for the reply JohnRC

A few possibilities suggest themselves:
1) activated charcoal as commercially presented might have some aromatic type impurities. Did you run a control blank experiment to see whether benzene was present when you simply exposed your (fresh as presented) adsorbent to the ethanol solvent you were using?

I did not do that, I will give it a try.

Did you try running the adsorption experiment in activated charcoal that had been pre-leached with the ethanol solvent you were using?

Do you mean, using the the carbon after it has been placed in the ethanol solvent (no benzene) into a solution which has benzene for the adsorption?

2) Ethanol is hardly the ideal solvent to use in an experiment like this, since ethanol is also adsorbed to some extent onto the active sites of a charcoal adsorbent. Could your results be explained if the adsorption of ethanol were actually stronger than that of benzene? Could you run your experiment with water as solvent? I know that benzene is only sparingly soluble in water, but it is more soluble than a typical aliphatic hydrocarbon, and low solubility is generally a good thing in adsorption studies. Note that you do not have to use water as the GLC carrier; the chromatography step of your experiment is essentially distinct from the adsorption step, and you could do the chromatography by starting with pure water and ramping up an admixture of ethanol.

I realize ethanol is not ideal (nor is methanol) but I have very few chemicals to choose from as my solvent. Ideally I would have used water but there are several problems as you already noted. First is the solubility (I also have to run toluene and xylene as my analytes which are also very insoluble in water) which possibly can be managed if I make larger solution volumes to be able to dissolve benzene in water.

The other is that I can't use it for GC, at least not for the GC setup I have. I don't have any more advanced sampling equipment like a headspace sampler or solid extraction. Plus the GC is equipped with a HP-5 capillary column which would definitely get damaged. You mention "ramping up an admixture of ethanol" can you please explain what do you mean by this?

 

[JohnRC]

Thanks for the reply JohnRC
*** (snip) ***
Do you mean, using the the carbon after it has been placed in the ethanol solvent (no benzene) into a solution which has benzene for the adsorption?

Yes that is exactly what I did mean.

I realize ethanol is not ideal (nor is methanol) but I have very few chemicals to choose from as my solvent. Ideally I would have used water but there are several problems as you already noted. First is the solubility (I also have to run toluene and xylene as my analytes which are also very insoluble in water) which possibly can be managed if I make larger solution volumes to be able to dissolve benzene in water.

The other is that I can't use it for GC, at least not for the GC setup I have. I don't have any more advanced sampling equipment like a headspace sampler or solid extraction. Plus the GC is equipped with a HP-5 capillary column which would definitely get damaged. You mention "ramping up an admixture of ethanol" can you please explain what do you mean by this?

OK -- I quite see, and take your point about the unsuitability of water as a GC solvent; I guess I was thinking more about column liquid chromatography, which you can often run with a carrier solvent that starts out as something in which your intended analyte is quite insoluble and change the carrier solvent by steadily increasing the proportion of a solvent in which the analyte is more soluble.

For your system, there might be a practical alternative in something like an abstraction of benzene/toluene/xylene into a solvent like chloroform before injection onto the GC.

That is a sequence like (1) stirred adsorption equilibrium in aqueous solution (2) filter off adsorbent (3) solvent extraction of filtrate with chloroform, and layer separation (4) GC run on chloroform layer.

I am not at all confident that I have the answer to your problems, but thought that throwing some ideas into the ring might be useful to you.

 

[FaNgS]

For your system, there might be a practical alternative in something like an abstraction of benzene/toluene/xylene into a solvent like chloroform before injection onto the GC.

That is a sequence like (1) stirred adsorption equilibrium in aqueous solution (2) filter off adsorbent (3) solvent extraction of filtrate with chloroform, and layer separation (4) GC run on chloroform layer.

I am not at all confident that I have the answer to your problems, but thought that throwing some ideas into the ring might be useful to you.

This is more than excellent. Thank you for the great idea.

I've done liquid liquid extraction a long time ago and it's a bit fuzzy in my head. Just a few questions in regards to it. Say I use water as my solvent to prepare the benzene solution for adsorption and then use something like hexane as the extractor solvent (name?).

Would the benzene (or toluene or xylene) transfer completely from water to hexane?

Also for the calibration curve using internal standard, it would be the same except I would use ethyl acetate in hexane?

 

[JohnRC]

This is more than excellent. Thank you for the great idea.

I've done liquid liquid extraction a long time ago and it's a bit fuzzy in my head. Just a few questions in regards to it. Say I use water as my solvent to prepare the benzene solution for adsorption and then use something like hexane as the extractor solvent (name?).

Would the benzene (or toluene or xylene) transfer completely from water to hexane?

Also for the calibration curve using internal standard, it would be the same except I would use ethyl acetate in hexane?

With liquid/liquid extraction, you get a partitioning constant -- at equilibrium, the activity (concentration) of the analyte in the two solvents is in a constant ratio (partition coefficient) at a particular temperature.

For example a(Iodine in chloroform at 25°C)/a(iodine in water at 25°C) = 70

Points to watch out for:
• the partition ratio can be very sensitive to temperature.
• It might be hard to find the partition ratio tabulated. At worst, the ratio of solubilities will give you a general guide (would be perfect if the solutions were ideal). But you should not need to know the partition ratio. All you really have to do is to ensure that a fixed percentage of whatever benzene concentration comes out of the adsorption step is presented at the GC inlet. So measured volumes are crucial.
• Hexane is usually regarded as a fairly poor solvent; it is less dense than water, and it spreads on the surface of water. Chloroform is a better solvent, forms a more definite interface, and is more dense than water. On the other hand, water is more soluble in chloroform than hexane, and a typical preparative technique like drying the solvent with anhydrous calcium chloride is not likely to be viable in an analytical procedure. So try with spectroscopic grade hexane (solvent grade likely to contain aromatic impurities)

Because of the need for careful volume management, I think that the best place to introduce the internal standard is just after the filtering, before the liquid-liquid extraction. That will ensure that small errors in volume management or evaporation from that point onward will not carry through to analytical results. But a complex analytical procedure like the one proposed will always require fairly extensive testing, validation, and calibration.

 

[FaNgS]

Thank you very much JohnRC. Thanks to you I was able to determine that the issue was with my carbons and my equipment is also to blame.

But I still have some queries regarding adsorption experiments.

Since I'm still at the early stages of my experiments, I first must determine the time it takes for my activated carbons to reach equilibrium. What is the proper way to do this?And how do I know when the equilibrium is achieved?

I've conducted an experiment where I analyzed the concentration of samples over a period of three days in 24 hour increments. But at 48 hours the drop from the initial concentration was less than what was observed at 24 hours.

Experimental Data: I have 3 separate vials with initial concentration solutions 400 mg/L

At 24 hours after adsorption the concentration dropped to 378 mg/L while after 48 hours it dropped to 380 mg/L and at 72hours it dropped to 392 mg/L.

 

[JohnRC]

Thank you very much JohnRC. Thanks to you I was able to determine that the issue was with my carbons and my equipment is also to blame.

But I still have some queries regarding adsorption experiments.

Since I'm still at the early stages of my experiments, I first must determine the time it takes for my activated carbons to reach equilibrium. What is the proper way to do this?And how do I know when the equilibrium is achieved?

I've conducted an experiment where I analyzed the concentration of samples over a period of three days in 24 hour increments. But at 48 hours the drop from the initial concentration was less than what was observed at 24 hours.

Experimental Data: I have 3 separate vials with initial concentration solutions 400 mg/L

At 24 hours after adsorption the concentration dropped to 378 mg/L while after 48 hours it dropped to 380 mg/L and at 72hours it dropped to 392 mg/L.

Experimentation in this area can be enormously difficult. In the ideal situation adsorption should simply increase with time, and the equilibrium will be a practical compromise where the upward trend in adsorption (downward trend in residue analysis) falls below an estimated error scatter. So why, in your case, has the adsorption fallen with increasing time?

Let us first establish a few things, because I do not know the full detail of your current procedure.

(1) Are you still using an alcohol solvent? One possible explanation of falling adsorption with time is that an alcohol is undergoing slow atmospheric oxidation to the corresponding fatty acid, which is strongly adsorbed, and displacing your analyte. As a more general form of this suggestion, is there any competing substance that may be present either as the result of a chemical reaction, or of diffusion in from elsewhere, or that simply has a slower adsorption profile due to an activation energy barrier to adsorption?

(2) I note that you did the 1/2/3 day experiments with different vials. Were the masses of adsorbent identical? Did the adsorbent come from the same batch? I would have said of the 378 and 380 results that they were the same within experimental error, but 392 quite clearly is not!

(3) Is your activated carbon in powder or pellet form? Pellets are much easier to handle, and in particular they should allow decanting rather than filtering -- provided that they are dust-free. With either pellets or powder there is a question of passage of fine particles of adsorbent (and the analyte they are carrying) through whatever filtration/decanting procedure you are using.

(4) Were the adsorption vials thermostatted throughout? A weak adsorption can be extremely temperature sensitive, and if the vials were to follow daytime/nighttime temperature changes, then your analyses could be anywhere! There would be active adsorption during the cooler night hours and desorption during the warmer day hours.

(5) The general nature of your results suggests a very weak adsorption, and that is what might be expected for benzene on charcoal anyway. There are obvious reasons why most of the meaningful adsorption studies are done with systems with moderate adsorption strength. Very strong adsorption means that any analyte that encounters a free surface site will be taken up; very weak adsorption means that all sorts of competing factors and side-issues come into play.

 

[FaNgS]

Thank you again JohnRC. I'll answer your questions below in blue.

Experimentation in this area can be enormously difficult. In the ideal situation adsorption should simply increase with time, and the equilibrium will be a practical compromise where the upward trend in adsorption (downward trend in residue analysis) falls below an estimated error scatter. So why, in your case, has the adsorption fallen with increasing time?

Let us first establish a few things, because I do not know the full detail of your current procedure.

(1) Are you still using an alcohol solvent? One possible explanation of falling adsorption with time is that an alcohol is undergoing slow atmospheric oxidation to the corresponding fatty acid, which is strongly adsorbed, and displacing your analyte. As a more general form of this suggestion, is there any competing substance that may be present either as the result of a chemical reaction, or of diffusion in from elsewhere, or that simply has a slower adsorption profile due to an activation energy barrier to adsorption?

Yes, I'm using an alcohol solvent. I'm rather restricted in solvent choice. I'm using ethanol absolute which is 99.5% pure. As mentioned previously I would ideally use water but the GC isn't equipped to handle water. I've tried liquid-liquid extraction but controlling the separation and quantities was rather difficult and time consuming.

I don't believe that there is any other substance competing for adsorption since all the chemicals (solvent and analyte) I'm using have a purity of 99% or higher. I wouldn't believe there would be any chemical reaction taking place either since the compounds are stable. As for diffusion this is possible. But I've been using blank solutions without carbon as controls and used them to account for any diffusion losses. As far as I've seen in literature the adsorption process taking place is exothermic and I've also maintained the experiment in a water bath at 30 deg C throughout the 72 hours.

Is it possible that the ethanol is displacing the benzene after the carbon gets saturated with benzene??

(2) I note that you did the 1/2/3 day experiments with different vials. Were the masses of adsorbent identical? Did the adsorbent come from the same batch? I would have said of the 378 and 380 results that they were the same within experimental error, but 392 quite clearly is not!

The masses were somewhat identical. I tried to maintain the amount of carbon to be 0.050g so the variation in amount was 0.049g to 0.053g. I dried some quantity of each carbon for 24 hours at 110 deg C and after drying I measured out 0.050g and displaced them in the vials.

(3) Is your activated carbon in powder or pellet form? Pellets are much easier to handle, and in particular they should allow decanting rather than filtering -- provided that they are dust-free. With either pellets or powder there is a question of passage of fine particles of adsorbent (and the analyte they are carrying) through whatever filtration/decanting procedure you are using.

They are in pellet form but I crushed them into powder using a mortar and pestle then dried as in (2).

Ideally, I should be using placing my experiment vials in a water bath and shaker to keep the carbon suspended and continuously mixing in the solutions but there isn't one available so I'm using a water bath. The carbon would settle at the bottom of flask and I would filter out the solution although I can't see any particles suspended in the filtrate. I then dispense the filtrate into the GC vials using a micro-pipette. I also tried centrifuging the filtrate for 5-10 minutes but there wasn't any particles that were sticking on the walls of the centrifuge vial. So I've been only filtering and canceled the centrifuging step.

(4) Were the adsorption vials thermostatted throughout? A weak adsorption can be extremely temperature sensitive, and if the vials were to follow daytime/nighttime temperature changes, then your analyses could be anywhere! There would be active adsorption during the cooler night hours and desorption during the warmer day hours.

Yes. I've maintained all the experiment vials in a thermostat controlled water bath at 30 degrees Celsius throughout the 72 hours. I ensured that the temperature was 30 degrees Celsius throughout the 72 days using a separate thermometer.

(5) The general nature of your results suggests a very weak adsorption, and that is what might be expected for benzene on charcoal anyway. There are obvious reasons why most of the meaningful adsorption studies are done with systems with moderate adsorption strength. Very strong adsorption means that any analyte that encounters a free surface site will be taken up; very weak adsorption means that all sorts of competing factors and side-issues come into play.

Several factors play part in the adsorption process even the adsorbent itself. Literature presents several experiments for the adsorption of benzene using commercial and carbon made from biomass such as coconut shells. But I can't really compare or relate much since they mostly use water as their solvent or do their experiments in the gaseous phase. I haven't been able to find a paper whose authors used alcohol as their solvent in their experiments.