Grinding

10 Dec 2002
Dear Steve and others
I was looking for grinding and milling equipments in the Exhibit Room on GSA Denver 2002. There was a Booth #421 exhibiting this type of equipment. Here is the information: ----------------------------------
SPEX CertiPrep 6750 FREEZER/MILL The 6750 Freezer/Mill from SPEX CertiPrep is an updated version of the classic SPEX CertiPrep 6700 cryogenic mill. The 6750 is used on samples normally considered difficult or impossible to grind at ambient temperatures: polymers, wood, bone, tissue, rubber, and the like. The Freezer/Mill embrittles samples by immersing the grinding vial in liquid nitrogen, and then pulverizes them with a magnetically driven impactor. The samples, up to several grams in weight, are isolated in a sealed vial during grinding, which rarely takes more than several minutes. Grinding vials can be pre-cooled to achieve throughput of 10 or more samples per hour; low temperatures are retained throughout, preserving structural and compositional aspects degraded or volatilized during room-temperature grinding.

Telephone: 1-800-LAB-SPEX or (732) 549-7144
Fax: (732) 603-9647
Email: CertiPrep@spexcsp.com
http://www.spexcsp.com

Since I didn't have chance try that type of mill on the show, I don't know its performance. But I saw some plant samples they ground on exhibition. The sample grain size looks meeting my requirement for cellulose extraction. The Exhibitor told this equipment does make big noise.
--------------------------
Zheng-Hua
Dr Zheng-Hua Li
Stable Isotope Lab
Dept of Geological Sciences
Uni of Tennessee
Knoxville, TN 37996

____________________________________________________________________________________________________________

10 Dec 2002
Had to try this again since pics wouldn't go to list:

Steve & group,

We use two refurbished Erbach (E-300?) ORBITAL shakers. Recently picked up a 1960 New Brunswick G-10 model too. Vintage. All three we have were purchased through university surplus. New price for these heavy-duty models is over $3k! Make sure their not reciprocal shakers if you find used ones.

The top plate has been modified (differently on each one) to hold Gerberbaby food jars. These are expendible, cheap, readily available and have nice curved inner edges for ball/jar surface contact. IMPORTANT: We wrap a strip of packing tape around the lower half and base of the jars to retain sample material if it breaks.

One shaker has a heavy foam block (see pics; plastic bags you see in 4th pic are baggies used before the packing tape idea) on the top platform withdoorknob holes drilled in for jars. The other two use the standardErlenmeyer flask metal clips that came with them originally, which have been bent a bit to fit. Inside jars we put 2 stainless-steel balls (0.5" and 0.75" diam). Lids get label tape to id. The amount of material you add should just cover the floor of the jar. Too little leaves a mound in the center unground; too much inhibits ball movement.

We run them at low speed only; overnight for most dry leaf/needles. Wood must be 'thoroughly' preground on a Wiley mill before going on the shaker. Wood usually needs several days grinding. Mortar and pestle with liquidnitrogen are kept handy for the ones that are insufficiently ground (still this is tough).

Overall, this works great for dry plant material. Somewhat less effective for wood. The Wiley pregrinding is key for wood but has high material loss.

Bob
--------------------------------------------------------------------------
Robert D. Stickrod II, Lab Manager Idaho Stable Isotopes Laboratory
University of Idaho, Forest Resources CNR 214, 6th & Line St. Moscow, ID
83844-1133, USA (208) 885-6512 lab (208) 885-6226 fax (509) 338-2814 cell
www.cnr.uidaho.edu/isil
--------------------------------------------------------------------------

___________________________________________________________________________________________________________

11 Dec 2002
Hi Steve and other tree ringers

Todd Dawson turned me on to a grinding method that I have found to be quite good and cheap. Essentially we are using a ball mill called a Wig-L-Bug. It is made by Crescent (industrial model C32003A) and can be found from dental supply houses (Dentsply) and costs only about $370. I would highly recommend purchasing spare "long arms" as they tend to break periodically (only $3 each). It comes with a simple timer and anywhere from 5 to 10 minutes will produce powdered wood samples.

We use 3/16 inch ball bearings (very cheap) that we clean will a sonic bath. We use disposable 1 ml cryotubes (NUNC Brand #375299, about $160 for a pack of 500) that fit well onto the Wig-L-Bug. The cryotubes appear sturdy enough for the shaking but occasionally the seam may crack (we add some tape to the end and have never lost a sample). The other thing I like is that the cryotube mates perfectly with a 1.5 ml epindorph tube (which is what I store my samples in) so transferring material into and out of the tube is pretty easy. The tube can hold up to 100 mg of wood at a time. The other advantage of a ball mill is that the sample is homogenized as well as ground, so subsampling is less of an issue. If you have multiple ball mills and fairly soft wood at least 20 to 30 samples could be ground per hour.

I have yet to process very small samples and getting enough of the powder out of the cryotube could be problematic. I have found that about 4 to 5 mg of powder cannot be extracted from the cryotube. Most of my samples are about 60 to 80 mg so the loss is acceptable at this point. The losses are better, in my opinion, than those from a Wiley mill. To avoid losses in grinding, I have contemplated for very narrow tree rings (< 0.5 mm), cutting the ring portion of interest into very narrow slices and placing them directly into the filter bags for processing and then using a blade to cut up the cellulose for loading into capsules. Has anyone done this and does the cellulose extraction procedure have to run longer to process slices?

To optimize Wig-L-Bug usage the wood needs to be completely dry and cut into small pieces (large chunks will take much longer). Generally, I cut out the latewood or earlywood under a dissecting scope with a scalpel and then take the slices and cut them into small pieces with a razor blade, place them in an epindorph tube, dry them overnight and then grind them the next day. I am using the filter bag/soxhlet method and there is some concern that the powder is so fine that it will pass through the pores of the filter bag. However, I have found that I get comparable yields from the powder as compared to coarse samples (25-27% compared to 27-33%). The reduction in extraction time for powdered samples as compared to coarse samples makes these losses acceptable, but also something to consider when dealing with small amounts of material (again, I may not grind very narrow slices). The fineness of the powder is a function of the time in the ball mill, so some tests should be run to determine what works best. Hardwoods or latewood would obviously take longer to grind than softwoods or earlywood (helps to place very fine pieces of hardwood in the mill, whereas softer woods could have larger chunks). That's all I can think of.
Happy Grinding and Merry Christmas to all!

John Roden
Associate Professor
Department of Biology
Southern Oregon University
Ashland, OR, 97520
rodenj@sou.edu
voice: 541-552-6798
fax: 541-552-6415

____________________________________________________________________________________________________________

12 Dec 2002
Dear John and others,
> To avoid losses in grinding, I have contemplated for very narrow tree rings (< 0.5 mm), cutting the ring portion of interest into very narrow slices and placing them directly into the filter bags for processing and then using a blade to cut up the cellulose for loading into capsules. Has anyone done this and does the cellulose extraction procedure have to run longer to process slices?

As we are working with small wood samples (10-30 mg), we tested possibility to perform cellulose extraction on coarse samples (0.1-0.3mm thick slices and wood shreds) in order to avoid large losses due to grinding (we are using the Retsch Mixer Mill MM200 with 10 ml stainless steel jars). After a 2-day bleaching step, these samples (Jack pine) appear perfectly white, just as the ground ones. No isotopes analysis have been performed on them yet.

> I am using the filter bag/soxhlet method and there is some concern that the powder is so fine that it will pass through the pores of the filter bag. However, I have found that I get comparable yields from the powder as compared to coarse samples (25-27% compared to 27-33%). The reduction in extraction time for powdered samples as compared to coarse samples makes these losses acceptable, but also something to consider when dealing with small amounts of material (again, I may not grind very narrow slices). The fineness of the powder is a function of the time in the ball mill, so some tests should be run to

We obtained a very fine powder with the MM200, but losses can be too important with our smallest samples as 5 mg in average stay in the jars (which sometimes correspond to 50% of our sample). Actually, grinding yield depends largely on the wood quality (i.e. large inter-specific variations, but no evident difference between sapwood and heartwood for coniferous species).

After some tests, we abandoned the idea to use F57 Ankom filters because of their too large porosity (30 microns) causing important yield decrease (especially during the step of removing of hemicelluloses by NaOH). I used to use teflon pouches (Alltech Pre-cut membranes 1 um PTFE 47mm 100/CAN $127.00 Part.2059). These filters are very resistant and can be used several times after careful washing. However, we recently had to switch to the 0.45 micron model (Part. 2058) because of serious delivery time problems with the supplier for the 0.1 um one. This 0.45 um model seems to be suitable too, but we don't really know what should be the minimum pore size to allow good diffusion of the extracts (i.e. lignins and hemicellulose fragments). Our alpha-cellulose yield is quite high (between 25% and 37%, according to the species), but we don't know if it is due to very small losses or mediocre hemicelluloses extraction.

Regards,
Stephane Ponton

Stéphane Ponton, PhD
Department of Biological Sciences
University of Lethbridge
4401 University Drive
Lethbridge, Alberta T1K 3M4, Canada

____________________________________________________________________________________________________________

28 Jan 2003
Hello,
Here is an addition to the previous discussion about sample reduction (grinding/maceration/etc.). I have been looking for equipment that will reduce small whole wood samples (30mg-50mg) to a particle size of less than 150 microns in a single procedure, with very small sample loss. The size constraint is based on a paper by Borella et al. (1998).
Borella, S., M. Leuenberger, M. Saurer, and R. Siegwolf, Controlling uncertainties in d13C analysis of tree rings: pooling, mixing and cellulose extraction, Journal of Geophysical Research, 103 (D16), 19519-19526, 1998.
So far I have two trials to report:
Gilson Company, Fritsch LC-140 Rotor-Speed Mill (called "pulverisette 14" by Fritsch)
I arranged to have samples of spruce ground at Gilson Company, Lewis Center, Ohio. This model was recommended to me by the technicians at Gilson after hearing about the desired particle size and beginning sample size constraints. The spruce I sent them was pre-ground to 20 mesh (840 microns).
I wanted the material ground in a single step, but my instructions were apparently not clear, as they tested the grinding in a 3-step procedure as follows:
(a) Initial grinding on 0.20mm sieve (200 microns); 20% retained on a 106 micron sieve, 80% passed through.
(b) Half of the material was then ground on a 0.08mm (800 micron) sieve. After grinding, almost all material passed through the 106
micron sieve, about 14% was retained on a 75 micron sieve, and about 85% passed through the 75 micron sieve.
(c) As a final test, they used 5mg samples from (b), and tested the equipment for amount recovered. At 20,000 rpm, about 80%
was recovered. At 10,000 rpm, 100% of the 5mg was recovered.
According to the technician at Gilson, cleaning the mill between samples would take "a matter of minutes". From the cleaning description I would estimate 5+ minutes. I am told that underneath the rotor there are some ringed grooves. Material will accumulate in these grooves, and should be cleaned out using a wire brush. (This suggests one place where sample would be lost.) Another potential problem is buildup of static electricity on the metal parts. (Static electricity is a particular problem where wood particles are concerned.) The technician said that they sometimes use micronized silica to reduce the static, but admitted that it doesn't always work.
In subsequent email with the technician I was told that grinding a 50mg sample from 20 mesh (840microns) to <150microns would take about 10 seconds. Also, a titanium conversion kit (expensive) is available for "iron-free" grinding.
In summary, this equipment appears to be more than adequate for the grinding of whole wood to an acceptable particle size and the grinding is very fast. The cost of this mill is $5645. The accessories used for these tests were: (1) a 24-rib stainless steel rotor, $1340, (2) a 20mm sieve ring, $131, and (3) an 8mm sieve ring, $131. Only one sieve ring would be required, so the total cost to be operational would be $7116.
The problems I see with this system are sample loss caused by a non-sealed system and by the effects of static electricity, and between-sample "downtime"..
The URL for this company is
www.globalgilson.com
SPEX certiprep, 6750 Freezer Mill
I visited SPEX certiprep, in Metuchen, New Jersey, twice to test their 6750 Freezer Mill. This model is an impact mill, using electromagnets to shuttle a steel rod from one end to the other of steel cylinders, fitted with end caps. The setup uses liquid nitrogen to both cool the electromagnets and the sample(s). This mill can run 3 small samples at once using microvials, though the microvials use smaller rods, so the force of the impact is less. The sample capacity of the microvials is 0.1-0.5ml. Larger vials can accomodate samples of 0.5-4.0ml, but only 1 sample can be ground at a time.
The recommended grinding procedure is a 15-minute initial cooldown (this would be less for subsequent samples), and 3 2-minute impact sessions separated by 2 2-minute cooldown periods. I did not experiment to see if the same results could be obtained if these times were reduced. Given the specifications above, the per-sample grinding time using the microvials is about 5 minutes (assuming an average of 5 minutes for the cooldown).
I tested this mill with (1) the same spruce sent to Gilson, (2) thick slices from a 5.0mm dried conifer core, (3) thick slices from a 5.0mm fresh conifer core, and (4) approx. 3-8mm segments of fresh conifer needles. In almost all cases, the Freezer Mill was able to reduce the material to much less than 150microns. This is a somewhat subjective assessment, as I did not have the equipment necessary for a particle analysis, but the final product was much finer than material described as <106microns from the trial at Gilson. The only failure of the equipment occurred when I placed over 100mg of thick increment-core slices into one of the microvials.
Reduction of that sample size to 50mg resulted in a very fine final particle size.
Because the grinding system is closed, the percentage of sample recovered depended entirely on operator error in opening the vials, not loss during grinding. The samples were easier to recover if the vials were warmed to room temperature before the vials were opened. The microvials were not always easy to open, sometimes requiring two pairs of pliers, or similar, to remove the endcaps. The endcaps are machined as press fit, so a tight fit is desirable, but the company needs to come up with a simpler design for opening the vials. With care, one could recover well over 90% of any sample.
I used the vials they had been experimenting with in their testing lab. I did not carefully clean their vials before use, and they had been run for a long time without samples in them immediately prior to our use. The inside of older vials show some wear from the impactors hitting the sides of the cylinder. The samples showed some discoloration after the first run; probably caused by contact of the sample material with the inside of the steel vials. Subsequent uses of the vials did not show significant
discoloration, but some material from the inside of the vials was probably present. Cellulose extraction should remove this contaminant, but it may be a concern if whole plant material is to be analyzed.
I recommend this mill for the grinding of small botanical samples, with two caveats. The mill is very loud when running and a ready supply of liquid nitrogen must be available during milling. Also, sieves are not part of this method, so production of samples of a certain range of particle sizes is not possible using this equipment.
The cost of this mill is $4100, with large vials being $165, and a set of 3 small vials being $465. I would recommend at least 3 of each type of vial or vial set, to reduce the downtime during processing, thereby reducing both the total processing time and the liquid nitrogen cost. The total cost to be operational for small samples would be $4565, though additional sets of vials are recommended.
The problems that I see with the SPEX Freezer Mill are:
(1) the Liquid Nitrogen cost, and (2) sample loss. The vials are a press fitting, so you have to be very careful opening the vials, or you lose sample. We weighed the samples before and after, and found about a 10% loss. However, the samples were not dried before grinding, so it's possible that water was pulled out of the wood during the cooldown before grinding. Also, additional care in sample recovery might have increased the yield.
There was also a discoloration in the final product, apparently from the steel of the vials. We did not carefully clean their vials before use, and they had been run for a long time without samples in them immediately prior to our use. I don't have evidence that any powder from the interaction of the grinding rod and the vial would be removed during processing to cellulose, but I expect that this would be the case.
The URL for this company is
http://www.spexcsp.com/

/ / / /:/ / / / / / / / / /./ / / / / / / / / /./ / / / / / / / / /./
Dr. William E. Wright
Postdoctoral Fellow
Lamont-Doherty Earth Observatory
of Columbia University
61 Route 9W
Palisades, New York, USA
10964-1000
Phone: (845) 365-8441
FAX: (845) 365-8510
/ / / /:/ / / / / / / / / /./ / / / / / / / / /./ / / / / / / / / /./

___________________________________________________________________________________________________________

18 Jan 2004
A while ago I reported homogenization results for whole wood maceration using a cryogenic impact grinder built by SPEX CertiPrep. While I stand by my statements about the small particle size, and, therefore, the high degree of homogeneity in the resulting product, I must issue a caveat. I later found that grinding with this machine could produce wood flour with such a small particle size that much if not all of the test sample was lost through the pouching matrix during extraction. After experimenting with the grinding time, I was able to ensure that most of the particle sizes in the resulting product were between about 25 microns (the mesh size of the pouching material) and about 125 microns (e.g. Borella et al., 1998). (The upper end was not quantified. It was an estimate based on comparison of the post-grinding product with material passed through a 120 mesh sieve... 120 mesh=125 microns) (note: the amount of cellulose recovered after extraction plotted as a linear function against the time of grinding.)

That said, an internal cellulose standard produced from whole wood, using the SPEX equipment, currently has a standard deviation of less than 0.1 per mil for d18O.

Regards,
Ed Wright

--
/ / / /:/ / / / / / / / / /./ / / / / / / / / /./ / / / / / / / / /./
Dr. William E. Wright
Postdoctoral Fellow
14A Marine Biology
Lamont-Doherty Earth Observatory
of Columbia University
61 Route 9W
Palisades, New York, USA
10964-1000
Phone: (845) 365-8441
FAX: (845) 365-8150
/ / / /:/ / / / / / / / / /./ / / / / / / / / /./ / / / / / / / / /./

______________________________________________________________________________________________________________