New Techniques for Si Analysis Part 1

It has been a long time since my last substantial post, for which I apologize. My onkly excuse is that I have been extremely busy with classwork, research, and several side project (which I will write about at some later date). As far as my research goes, I do have some updates to report regarding the Si extraction and analysis procedure.

1. Traditionally I have been using ~0.075 g of soil coupled with 100 ml 0.2M NaOH. I have been experimenting with the same amount of soil, but a much larger volume of NaOH (250 or 500 ml). This allows me to take a much larger aliquot, which lessens problems due to pipetting errors. Interestingly, I have been finding higher values of Si when I use a larger volume of NaOH (see graph). There may be several reasons for this: a) there may be residual Si on the inside of the bottles, the total amount of which may be magnified simply because of the larger aliquot taken; b) the larger aliquot (and NaOH volume) may be causing some weird absorption probem. I have observed in the past that NaOH, when mixed with molybdate before it is totally neutralized by acid, can raise absorption values by ~0.01. This is substantial.

The larger aliquot extracts probably require their own unique standard curve. If the NaOH is causing increased absorbance, or if there is residual Si, then it should be possible to correct for this in the standard curve in two ways. First, by premixing the Si standard with 0.2M NaOH, it will mimic conditions present in the bottle. When an aliquot of the NaOH/Si is added to the reaction flask (and the molybdate and acid reagents), it will be added in exactly the same fashion as the extractions. (By the way, I am now adding the molybdate reagent before I add the Si aliquot (see below). This could be critical, as the NaOH may not be neutralized completely by he acid (especially a larger aliquot). Second, water used to make the NaOH/Si standard should be first stored in one of the reaction bottles, to account for residual Si. Using these two steps, I should be able to determine if the problems outlined above are the real culprits.

2. The amount of 0.5M H2SO4 added to the flask matters. In the past when I was taking small aliquots (in the range of 0.25 ml) and adding 5 ml of H2SO4, this may have been problematic. The key is to keep the pH below 1.5 in the flask, which 5 ml of H2SO4 does, but the ionic strength may have been too great. An ionic strength of 0.5 and above may cause problems with the molybdate reaction. To account for this, I have now adjusted the acid volume in the flask relative to the Si aliquot volume. For example, an Si aliquot of 0.25 ml receives 3.9 ml H2SO4, while an Si aliquot of 1 ml receives 5 ml H2SO4, and so on. In this way, the pH and ionic strength remain low.

3. The mixing order of the reagents matter. In the past I have added the acid to the flask, followed by the Si aliquot, and then the molybdate reagent. This is not a problem for the molybdate, but it is problematic for the Si aliquot. Something happens to the Dissolved Si when it is added to only acid. My guess is that it polymerizes with some other compound present (perhaps the NaOH?). When this happens, it will not be able to combine with the molybdate. As I mentioned earlier, I suspect that unneitralized NaOH can raise the absorbance. Thus, the NaOH may be reacting directly with the molybdate. To account for these problems, I have begun adding the molybdate reagent to the flask before the Si aliquot. This has greatly reduced scatter problems. One would think that adding the Si aliquot before the molybdate would be preferrable, as the NaOH would be neutralized. I have found that this isn't the case.

More soon...

Currently Waiting for Stats to Begin

So I figured I would blog about it; something I have not done since February. Shame on me.

So THAT'S How You Dissolve Silica!

My current Si dissolution procedure:

Extraction Steps 12 February 2008
Items outlined in red have not been attempted yet

Items needed
50 ml Nalgene bottles
100 ml Nalgene bottles
Chemicals outlined by Jones & Dreher
pH meter
5N NaOH
0.5 M H2SO4
Digital pipette
Cuvettes

Reagents: follow the procedures outlines by Jones & Dreher.

Dissolution procedure
Weigh and record a dry 50 ml Nalgene bottle (cap included).

Place approximately 0.38 g soil in the bottle. Weigh and record.

Allow the soil sample to dry at least 2 hours at 70˚C, then weigh and record. Set aside.

Determine plant available Si
Add approximately 50 ml Academic water to another 50 ml Nalgene bottle. Record the time.

Place in the pre-heated 85˚C water bath for at least 0.5 h.

Using the digital pipette, add 48 ml of the heated water to the soil sample bottle. Swirl the mixture gently. Tighten the cap and place it in the water bath for 1 min.

Remove a 1 ml aliquot using the digital pipette. Place the aliquot in a 25 ml (class A) volumetric flask which is pre-filled with 5 ml 0.5 M H2SO4. Flush the pipette tip into the flask with Academic water at least 2x to remove any residual Si. Record the time. Set aside. To determine the absorbance, go to the spectrophotometric procedures section.

Add 2 ml of 5N NaOH to the soil sample bottle. Replace and tighten the cap, and swirl the mixture. Loosen the cap approximately ¼ turn from tight. Place the bottle in the water bath, and turn the agitation speed to 4.5. Record the time.

Remove 1 ml aliquots at pre-determined times (10 min, 0.5 h, and so on up to 5 h). Record the extraction times. Depending on the extraction time, the concentration of Si in the aliquot will be up to 1000 µg. Since this is too high for the spectrophotometric procedures, the aliquot must be diluted.Place the aliquot in a 25 ml (class A) volumetric flask which is pre-filled with 2.5 ml 0.5M H2SO4. Flush the pipette tip into the flask with Academic water at least 2x to remove any residual Si. Bring the solution to mark (meniscus bottom should be at the line). Shake the flask to mix the solution. Transfer the diluted Si solution to a dry 50 ml Nalgene bottle.

At this point, the Si aliquot is diluted 25x. For example, if the original aliquot holds 500 µg Si, its concentration is 500 µg Si ml-1 H2O. When diluted, the flask still holds 500 µg Si, but now its concentration is 20 µg Si ml-1 H2O. However, this is still too high.

Extract 2.5 ml of the diluted Si solution with the digital pipette.

Place the diluted aliquot in another 25 ml (class A) volumetric flask which is pre-filled with 5.0 ml 0.5 M H2SO4. Flush the pipette tip into the flask with Academic water at least 2x to remove any residual Si.

At this point the Si solution is diluted another 10x. Using the scenario outlined above, the 2.5 ml second aliquot (which has a concentration of 20 µg Si ml-1 H2O) will hold 30 µg Si total. When this aliquot is diluted in the second volumetric flask to 25 ml, its new concentration will be 0.2 µg Si ml-1 H2O, well within the range of the spectrophotometric procedures.

Proceed to the spectrophotometric procedures section

Spectrophotometric procedures
Set the spectrophotometer λ to 810 nm. Set the background to 0.200. Allow the machine to warm up for at least 0.5 h.

Add 5 ml of the molybdate reagent to the reaction vessel (the 25 ml volumetric which holds the 2.5 ml Si aliquot along with 5 ml 0.5 M H2SO4). Swirl the mixture. Record the time. Allow the reaction to continue for 5 min.

Add 2.5 ml 20% tartaric acid. Swirl the mixture. Record the time. Allow the reaction to continue for 5 min.

Add 1 ml of the reducing solution. Bring to mark. Place the cap on the reaction vessel and shake the mixture gently. Transfer to a dry 50 ml Nalgene bottle. Record the time. Allow the reaction to continue for 15 min.

Fill a clean cuvette with Academic water and record its absorbance in the spectrophotometer. Empty the same cuvette and fill with the blue solution from the reaction vessel. Empty the solution from the cuvette and refill. Record the absorbance.

Yes, I am Still Alive

It has been a busy few months since my last posting. You know how it goes: I kept meaning to post to the blog, but could never find the time. Well, I still don't have the time, but oh well. Some things are more important that school work. Like procrastinating.

My last few blogs have been reviews of various articles related to silica dissolution methods. Not exactly great reading material, but nevertheless important for my research. Since then, I have continued reading plenty of articles in my attempt to perfect (or at least get a little better at) my own particular dissolution procedure. I have decided to pursue an 85 C water bath dissolution method similar to that outlined by Sauer et al. To that end, I purchased a used water bath on ebay. The nice thing about this particular water bath is that it is also an orbital shaker. Thus, my samples will not only be heated, but also will be agitated. This is a big step, and should greatly speed up dissolution. I have found that using a 0.2 M NaOH solution works best, and have thus far had good results.

I have run into several sticking points however. First is the issue of silica contamination from glassware. The use of glass pipettes, volumetric flasks, etc. can greatly influence the amount of silica actually in the sample. This is especially bothersome since I am working with dissolved silica amounts in the neighborhood of 10 micrograms (0.0000010 gram). Any contamination from glass can greatly influence this. To get around this problem, I have been using as little glassware as possilbe. For example, I have purchased a digital pipette which uses plastic pipette tips instead of glass. I am still forced to use glass volumetric flasks when I add reagents to the silica sample (this is a step used to "color" the silica, so its concentration can be determined). To mitigate contamination in this step, I never add the silica sample to a dry flask. Instead, I always make sure that there is water in the flask to dilute the silica before it comes into contact with the glassware. On top of this, the silica sample is not left in the flask any longer than is necessary. Undoubtedly, there is some silica contamination. But as long as I am consistent with my procedures (i.e. each sample spends the same amount of time in the flasks), any variation in the data should be mitigated.