Post on 18-Jan-2016
Malmö, ABM meeting May 2012
Publication of the ABM ringtest studies onCEC and exchangeable cations
Dohrmann, R. BGR/LBEG
Malmö, ABM meeting May 2012
2 papers were accepted by Clays and Clay Minerals:
Interlaboratory CEC and exchangeable cation study of bentonite buffer materials: I. Cu(II)-triethylenetetramine method
Interlaboratory CEC and exchangeable cation study of bentonite buffer materials: II. Alternative methods
Dohrmann, R., Genske, D., Karnland, O., Kaufhold, S., Kiviranta, L., Olsson, S., Plötze, M., Sandén, T., Sellin, P., Svensson, D., Valter, M.
BGR/LBEG, S&B Industrial Minerals, Clay Technology, B+Tech, ETH Zürich, SKB
Malmö, ABM meeting May 2012
Motivation and questions
• Large differences of results reported during 2010 meeting
• Lab exchange of data to clarify:– How large is scattering (precision)?– Which results are more plausible (accuracy)?
Malmö, ABM meeting May 2012
What is a real difference?TR-09-29 (LOT) Svensson, 2010 (ABM meeting)
Mg2+ analysis of the same ABM samples differed between different labs
Mg2+ meq/100 g lab A lab BCalcigel 8.7 14.6MX80 3.6 8.6Rockle 9.7 17.2DepCAN 15.2 24.9Ikosorb 16.6 26.6…
Mg2+ a few meq/100 g
larger at contact
Malmö, ABM meeting May 2012
I. Cu-trien method
Malmö, ABM meeting May 2012
5 labs participated, Cu-trien details
water buffer %/Cu-trien max possible elemental Lab . (M) (mL) (mL) (mL) bentonite COX adsorbed (max.) CEC (meq/100 g) analysisLab 1 0.015 10 25 - ≈ 60 % ≈ 145 ICPLab 2 0.015 20 50 - ≈ 60 % ≈ 145 ICPLab 3-1 (setup1) 0.01 10 50 - 83% 107 AASLab 3-2 (setup1) 0.01 20 40 - 42% 213 AASLab 4 0.02 5 44 1 130-140 500 58% 152 AASLab 5 0.01 10 50 - 80+120 200+300 50%+34% 177 (120 mg) ICP
Cu-trien
200400 ± 50
200
sample mass (mg)
200
Lab technique 1 time (min) technique 2 time (min) technique 3 time (min)Lab 1 hand-shaking 15 sonication 10 rocking platform 30Lab 2 vibrating table 120 vibrating table 30Lab 3-1 shaker 120Lab 3-2 shaker 120Lab 4 sonication 3 30 times end-over, manuallyLab 5 end-over-end shaking 120
before Cu-trien addition after Cu-trien addition
Malmö, ABM meeting May 2012
Exchange population, precision (1)
Ca2+ data inaccurate, but partly with good precision
Ca2+ data inaccurate, but partly with good precision
exchangeable cation MX80 COX Dep. CAN ASHApopulationNa+/CEC (VIS) (%) 70 24 27 73K+/CEC (VIS) (%) 2 16 2 1Mg2+/CEC (VIS) (%) 8 32 30 16Ca2+/CEC (VIS) (%) 36 170 66 24control sum (%) 116 242 125 114
MX80 COX Dep. CAN ASHA
Na+ 58.4 2.6 22.3 63.3standard deviation 2.1 0.4 1.6 2.3
K+ 1.8 1.7 1.6 0.5standard deviation 0.4 0.2 0.3 0.2
Mg2+ 6.9 3.4 24.3 13.8standard deviation 0.5 0.4 0.7 0.6
Ca2+ 29.9 18.5 54.2 20.6standard deviation 1.1 4.6 2.1 0.4
CEC (VIS) 83.9 10.9 82.2 86.5standard deviation 1.5 0.7 1.9 2.1
CEC (ICP/AAS) 84.8 12.1 82.6 86.3standard deviation 3.9 2.8 3.8 3.7
Malmö, ABM meeting May 2012
Precision (standard deviation)
bentonites
COX
0.0
1.0
2.0
3.0
4.0
5.0
stan
dard
dev
iatio
n (m
eq/1
00 g
)
NaK Mg
CE
C (
VIS
)
CE
C (
ICP
/AA
S)
Ca
Malmö, ABM meeting May 2012
Conclusions Cu-trien
• The overall quality of the returned CEC results using Cu-trien method was good
• Some outliers were detected
• Exchange population (cations) exceeds CEC largely
• The most important question what is a real difference? can be evaluated based on ‚precision data‘ now
Malmö, ABM meeting May 2012
II. Alternative methods
without COX
Malmö, ABM meeting May 2012
Details of alternative methods
Cu-trien'5xcalcite'
Lab (mL) n (mg)Lab 1 0.50 *) ⅓ of 25 3 500 ICPLab 2 0.15 *) ⅓ of 50 3 800±50 ICPLab 4 1.0 25 6 100 KjeldahlLab 5 0.01 50 1 400+600 ICP*) the total volume used was 25 mL respectively 50 mL, which was added in three portions of approximately equal volume
(M)elemental analysis
NH4Ac NH4Clrepetitions sample mass
Lab technique 1 time (min) technique 2 time (min) technique 3 time (min)
Lab 1 hand-shaking 10sonication (homogenizer probe)
1.5 rocking platform 30
Lab 2sonication (ultrasonic bath)
15 vibrating table 120
Lab 4 end-over-end shaking 1.) 1000; 2.-5.) short Lab 5 end-over-end shaking 120
dispersion after index cation addition
Malmö, ABM meeting May 2012
Exchange populationexchangeable cation MX80 Dep. CAN ASHApopulation
Na+/CEC (%) 69 27 67K+/CEC (%) 2 2 1Mg2+/CEC (%) 8 27 15Ca2+/CEC (%) (all labs) 26 49 20control sum (%) 105 105 103
Na+/CEC (%) 69 27 67K+/CEC (%) 2 2 1Mg2+/CEC (%) 8 27 15Ca2+/CEC (%) (all) 20Ca2+/CEC (%) (only lab 2 and 5) 24 46control sum (%) 103 102 103
only most plausible Ca 2+ results
results of all labs
Malmö, ABM meeting May 2012
Accuracy?
*: inflated by chloride-rich pore water; **: inflated by sulphate-rich pore water (gypsum dissolution); ***: questionable if inflated by dolomite dissolution.
MX80 Dep. CAN ASHA
Na+ 58.8 *22.4 *60.8standard deviation 1.4 1.2 3.9
K+ 1.7 1.7 0.5standard deviation 0.1 0.2 0.1
Mg2+ 6.7 */***22.8 *13.4standard deviation 0.8 1.5 0.8
Ca2+ **20.2 **/***38.8 **18.6standard deviation 1.6 1.4 0.8
CEC 84.6 83.9 90.8standard deviation 2.1 0.5 3.3
Malmö, ABM meeting May 2012
Conclusions alternative methods
• Precision of alternative methods is good
• Accuracy is partly not attainable, here more information than just CEC analyses are needed
• K+ and CEC results are (mostly) accurate
• Na+, Mg2+, and Ca2+ results partly inflated by chloride- or sulphate-rich pore water and soluble phases
Malmö, ABM meeting May 2012
Thank you
Malmö, ABM meeting May 2012
What is a real difference?
• If lab A and lab B use the same method:– result A result B: good precision (± small deviation)– result A result B (difference > precision): individual error(s) note: check with standard clay
• If lab A and lab B use different methods:– result A result B: indication for good precision– result A result B (difference > precision): option 1) individual error(s) option 2) systematic difference = operationally correct (both?) option 3) „complicated minerals“ such as zeolites (specific
adsorption) or vermiculites (slow / incomplete cation exchange)
Malmö, ABM meeting May 2012
What is an accurate/precise CEC result?
0
20
40
60
80
100
120
different CEC methods
CE
C (
meq
/100
g)
Reuverton (BGR/LBEG standard)
Ag
TU
(m
od
)
Am
mo
niu
m a
ceta
te
bar
ium
ch
lori
de
Cu
-tri
en
Co
-Hex
amin
e
Ag
TU
cal
cite
Cu
-tri
en c
alci
te
Co
Hex
cal
cite
Cu
-tri
en 5
xcal
cite
0
20
40
60
80
100
120
different CEC methods
CE
C (
meq
/100
g)
ammonium acetate, 0.5 g and 2 g
barium chloride, 0.5 g
barium chloride, 2 g
calcareous hectorite (CMS source clay)
Compensation of two sources of error, occasionally good agreement
Malmö, ABM meeting May 2012
Example for good precision, single lab (38 bentonites)
y = 1.00 x - 0.0
R2 = 0.994
0
10
20
30
40
50
60
70
0 10 20 30 40 50 60 70
Cu-trien5xCc 1. series
Cu
-tri
en
5x
Cc
re
pe
titi
on
Na y = 1.00 x + 0.0
R2 = 0.987
0.0
1.0
2.0
3.0
4.0
5.0
6.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0
Cu-trien5xCc 1. series
Cu
-tri
en
5x
Cc
re
pe
titi
on
Ky = 0.99 x - 0.1
R2 = 0.9997
0
10
20
30
40
50
0 10 20 30 40 50
Cu-trien5xCc 1. series
Cu
-tri
en
5x
Cc
re
pe
titi
on
Mg
y = 1.00 x - 0.5
R2 = 0.998
0
10
20
30
40
50
60
70
80
0 10 20 30 40 50 60 70 80
Cu-trien5xCc 1. series
Cu
-tri
en
5x
Cc
re
pe
titi
on
Cay = 0.98 x + 0.4
R2 = 0.979
0
20
40
60
80
100
120
0 20 40 60 80 100 120
Cu-trien5xCc 1. series
Cu
-tri
en
5x
Cc
re
pe
titi
on
sumy = 0.98 x + 0.9
R2 = 0.987
0
20
40
60
80
100
120
0 20 40 60 80 100 120
Cu-trien5xCc 1. series
Cu
-tri
en
5x
Cc
re
pe
titi
on
CEC
Malmö, ABM meeting May 2012
exch
ange
able
catio
n po
pula
tion
(%)
Na+ K+ Mg2+ Ca2+ control Ca2+ (*) control
MX80 Dep.CAN ASHA
sum sum (*)
exch
ange
able
catio
n po
pula
tion
(%)
Na+ K+ Mg2+ Ca2+ control Ca2+ (*) control
MX80 Dep.CAN ASHA
sum sum (*)