1
00:00:09,966 --> 00:00:11,866
Hi, I’m Todd Houlahan.

2
00:00:11,866 --> 00:00:14,233
Welcome to the seventh edition
of our portable XRF series

3
00:00:14,233 --> 00:00:17,233
on best practice application
of the technology.

4
00:00:17,233 --> 00:00:21,033
Today I’m joined by Ted Shields,
the Vanta product manager.

5
00:00:21,033 --> 00:00:21,900
Hi Ted.

6
00:00:21,900 --> 00:00:22,833
Hey Todd, how are you doing?

7
00:00:22,833 --> 00:00:23,733
Really good, and you?

8
00:00:23,733 --> 00:00:24,333
Great.

9
00:00:24,333 --> 00:00:27,400
Alright, today were going to talk
about sample preparation.

10
00:00:27,400 --> 00:00:29,400
So how do pXRF users figure out

11
00:00:29,400 --> 00:00:31,666
how much sample preparation
they need to do?

12
00:00:31,666 --> 00:00:33,266
The level of sample preparation

13
00:00:33,266 --> 00:00:37,300
depends on the objectives
of each individual user.

14
00:00:37,300 --> 00:00:40,466
So it could be no sample preparation.

15
00:00:40,466 --> 00:00:42,433
It could be full sample preparation

16
00:00:42,433 --> 00:00:43,966
and everything in between.

17
00:00:43,966 --> 00:00:48,733
Ultimately a customer needs
to test, challenge,

18
00:00:48,733 --> 00:00:52,066
and then decide on what level
of sample preparation

19
00:00:52,066 --> 00:00:53,966
will meet their objectives.

20
00:00:53,966 --> 00:00:55,266
And we’re also going
to show them some

21
00:00:55,266 --> 00:00:56,900
of this equipment
that we have behind us, right?

22
00:00:56,900 --> 00:00:58,400
Indeed, we’re going to start out with

23
00:00:58,400 --> 00:01:00,633
really basic levels of equipment

24
00:01:00,633 --> 00:01:03,900
right through to quite advanced
levels of sample prep.

25
00:01:03,900 --> 00:01:06,533
So with pXRF we’re
taking an established

26
00:01:06,533 --> 00:01:09,833
laboratory technique and
bringing it out into the field.

27
00:01:09,833 --> 00:01:12,566
Exactly, and when taking
it out into the field

28
00:01:12,566 --> 00:01:14,966
we encounter all sorts
of different samples

29
00:01:14,966 --> 00:01:17,933
so ultimately the quality
of the data will be

30
00:01:17,933 --> 00:01:20,600
dictated by the quality of the sample.

31
00:01:20,600 --> 00:01:22,966
So lets look at the
journey of our samples.

32
00:01:22,966 --> 00:01:24,733
First you need to collect the samples

33
00:01:24,733 --> 00:01:27,066
and with pXRF you can collect
a whole lot of samples

34
00:01:27,066 --> 00:01:28,700
and a lot of data points
and that gives you

35
00:01:28,700 --> 00:01:32,466
very fine-grained information
about your project.

36
00:01:32,466 --> 00:01:35,866
Typically then a few of those
are sent off to the laboratory

37
00:01:35,866 --> 00:01:39,600
and they get more detailed
information on a particular sample.

38
00:01:39,600 --> 00:01:42,033
The way they do that
is to dry the sample,

39
00:01:42,033 --> 00:01:44,166
then crush it, and then grind it.

40
00:01:44,166 --> 00:01:46,266
And they do this to make
the sample homogenous

41
00:01:46,266 --> 00:01:48,200
before they dissolve it in acid

42
00:01:48,200 --> 00:01:51,566
or make a fused bead
or a pressed pellet.

43
00:01:51,566 --> 00:01:54,866
So if your data quality
objectives include

44
00:01:54,866 --> 00:01:58,100
comparing to laboratory data,
you may want to consider

45
00:01:58,100 --> 00:02:01,533
doing the same sorts of things
that a laboratory does.

46
00:02:01,533 --> 00:02:04,400
So let’s start with stage one
at the lab, drying.

47
00:02:04,400 --> 00:02:06,800
How does moisture affect
portable XRF results?

48
00:02:06,800 --> 00:02:08,300
There are two main effects,

49
00:02:08,300 --> 00:02:11,166
the first of which is that
the moisture in the sample

50
00:02:11,166 --> 00:02:12,733
absorbs some of the x-rays

51
00:02:12,733 --> 00:02:14,100
that would normally get back
to our detector.

52
00:02:14,100 --> 00:02:17,333
So that makes us under
report on some elements.

53
00:02:17,333 --> 00:02:22,066
The other is that in the factory
we calibrate on dry samples

54
00:02:22,066 --> 00:02:25,300
and the dry samples weigh a little
bit less than the moist samples.

55
00:02:25,300 --> 00:02:28,566
So this also tends to makes us
under report on moist samples.

56
00:02:28,566 --> 00:02:30,633
Yeah, in my experience we always

57
00:02:30,633 --> 00:02:33,233
under report on moist samples, always.

58
00:02:33,233 --> 00:02:36,233
Yeah, that’s right and so we
then get asked two questions,

59
00:02:36,233 --> 00:02:38,433
one, how much moisture
can you tolerate?

60
00:02:38,433 --> 00:02:42,900
And the second is, if the
moisture level is consistent

61
00:02:42,900 --> 00:02:44,500
can we just go ahead
and correct for it?

62
00:02:44,500 --> 00:02:46,433
OK, how much can we tolerate?

63
00:02:46,433 --> 00:02:49,433
On the transition metals,
heavier metals

64
00:02:49,433 --> 00:02:51,700
about 5 to 10 per cent
moisture seems to have

65
00:02:51,700 --> 00:02:55,066
very little impact on
portable XRF results.

66
00:02:55,066 --> 00:02:58,966
And question 2 is a little bit
more complicated to answer.

67
00:02:58,966 --> 00:03:02,100
Different elements are
impacted in different ways

68
00:03:02,100 --> 00:03:04,633
by different levels of moisture.

69
00:03:04,633 --> 00:03:07,833
So, again, send the
samples to the lab.

70
00:03:07,833 --> 00:03:09,866
Get some moisture
measurements either

71
00:03:09,866 --> 00:03:13,800
before you send them using
a moisture meter or at the lab.

72
00:03:13,800 --> 00:03:15,233
Get the data back,

73
00:03:15,233 --> 00:03:18,266
have a look at the correlation
with your portable XRF results

74
00:03:18,266 --> 00:03:21,633
and see if the data
can be fit for purpose.

75
00:03:21,633 --> 00:03:24,866
Only the customer can decide
whether it’s OK or not.

76
00:03:24,866 --> 00:03:26,433
So do you have some
customer examples then?

77
00:03:26,433 --> 00:03:27,366
Yeah, lots!

78
00:03:27,366 --> 00:03:33,166
One example is customer
testing quite moist clay soils

79
00:03:33,166 --> 00:03:35,700
put all their samples in
clip lock plastic bags

80
00:03:35,700 --> 00:03:38,066
and then presses down on each one

81
00:03:38,066 --> 00:03:41,600
to sort of squeeze out as
much moisture as they can.

82
00:03:41,600 --> 00:03:44,966
Another example, one customer
putting all their samples

83
00:03:44,966 --> 00:03:49,233
on coffee filters and trying
to suck out as much moisture

84
00:03:49,233 --> 00:03:52,566
within a designated,
couple of hour period.

85
00:03:52,566 --> 00:03:54,900
Another customer puts
all their samples out

86
00:03:54,900 --> 00:03:58,333
for air drying for 24 hours
before testing them.

87
00:03:58,333 --> 00:04:00,600
So as long as you treat
your sample consistently

88
00:04:00,600 --> 00:04:02,700
you are likely to get
consistent results,

89
00:04:02,700 --> 00:04:04,466
probably really good for trending

90
00:04:04,466 --> 00:04:06,833
but to get the absolute
concentration value

91
00:04:06,833 --> 00:04:08,666
is probably, you need to evaluate.

92
00:04:08,666 --> 00:04:09,666
Exactly, yeah.

93
00:04:09,666 --> 00:04:14,366
One really interesting application
was testing the core cutter slurry.

94
00:04:14,366 --> 00:04:19,000
So collecting the slurry from
cutting say, a meter of core,

95
00:04:19,000 --> 00:04:21,466
drying it of course before analysis,

96
00:04:21,466 --> 00:04:23,466
but you’ve got a really fine powder

97
00:04:23,466 --> 00:04:25,533
representative of a meter of core,

98
00:04:25,533 --> 00:04:29,066
quite labor intensive but ingenious.

99
00:04:29,066 --> 00:04:30,500
Yeah so there are no rules.

100
00:04:30,500 --> 00:04:32,033
You can test whatever is at hand

101
00:04:32,033 --> 00:04:33,433
but then you need to evaluate it

102
00:04:33,433 --> 00:04:35,533
and make sure it really
is fit for purpose.

103
00:04:35,533 --> 00:04:37,233
Exactly, a bit of a common theme

104
00:04:37,233 --> 00:04:39,000
running through this series Ted.

105
00:04:39,000 --> 00:04:40,466
Indeed.

106
00:04:40,466 --> 00:04:41,933
There is also some good reading

107
00:04:41,933 --> 00:04:45,766
on the Association of
Applied Geochemists website.

108
00:04:45,766 --> 00:04:50,100
Work that Camiro
undertook in 2012-2013

109
00:04:50,100 --> 00:04:53,466
on all sorts of things but
moisture is included in there.

110
00:04:53,466 --> 00:04:55,600
Really good stuff.

111
00:04:55,600 --> 00:04:58,500
Read it in context
because the technology

112
00:04:58,500 --> 00:05:02,600
was six or seven years older
but worth a read, definitely.

113
00:05:02,600 --> 00:05:04,100
OK, moving on from moisture.

114
00:05:04,100 --> 00:05:05,866
Homogeneity.

115
00:05:05,866 --> 00:05:09,400
So there’s probably three
categories we can set up

116
00:05:09,400 --> 00:05:11,466
of different sample types.

117
00:05:11,466 --> 00:05:14,133
Samples requiring no preparation,

118
00:05:14,133 --> 00:05:16,166
samples requiring partial preparation,

119
00:05:16,166 --> 00:05:18,933
and samples requiring
full preparation,

120
00:05:18,933 --> 00:05:22,766
say down to 150-250 micron.

121
00:05:22,766 --> 00:05:26,966
So pXRF is used for so many
different types of samples,

122
00:05:26,966 --> 00:05:29,066
and for so many different uses

123
00:05:29,066 --> 00:05:30,933
it’s really hard to
give general advice,

124
00:05:30,933 --> 00:05:33,600
but in general what’s going on
is that you are trading off

125
00:05:33,600 --> 00:05:35,433
your precision for
how much time

126
00:05:35,433 --> 00:05:36,900
you want to invest in sample prep.

127
00:05:36,900 --> 00:05:39,000
Right, and so
because it’s so difficult

128
00:05:39,000 --> 00:05:43,300
to give generic advice let’s
use some real world examples

129
00:05:43,300 --> 00:05:44,900
for those three categories.

130
00:05:44,900 --> 00:05:47,566
Let’s start with
no sample preparation.

131
00:05:47,566 --> 00:05:49,966
So we have hundreds of customers

132
00:05:49,966 --> 00:05:53,933
testing directly on drill core,
directly on rock faces,

133
00:05:53,933 --> 00:05:59,300
directly on soils, directly on
un-prepared blast hole samples.

134
00:05:59,300 --> 00:06:02,300
And all of those
samples are heterogeneous

135
00:06:02,300 --> 00:06:06,566
so a lot of our customers are
taking multiple measurements

136
00:06:06,566 --> 00:06:09,933
and taking the average of those
readings to overcome that.

137
00:06:09,933 --> 00:06:13,166
They’re getting qualitative
semi-quantitative data

138
00:06:13,166 --> 00:06:14,733
but it’s fit for purpose.

139
00:06:14,733 --> 00:06:19,500
We’ve also got customers testing
gold in activated carbon,

140
00:06:19,500 --> 00:06:20,966
directly on the carbon.

141
00:06:20,966 --> 00:06:26,533
They’re also doing copper
in liquids in SXEW plants,

142
00:06:26,533 --> 00:06:28,533
no sample preparation.

143
00:06:28,533 --> 00:06:31,066
So there’s a wide
range of applications

144
00:06:31,066 --> 00:06:33,933
for customers doing
no sample prep.

145
00:06:33,933 --> 00:06:35,966
For the partial sample prep,

146
00:06:35,966 --> 00:06:38,000
those are typically done
for soil surveys

147
00:06:38,000 --> 00:06:41,233
which are doing really well for
things like base metals,

148
00:06:41,233 --> 00:06:44,800
for gold and pathfinder
elements for gold.

149
00:06:44,800 --> 00:06:46,633
That’s where we typically
find the partial

150
00:06:46,633 --> 00:06:47,833
sample prep being used.

151
00:06:47,833 --> 00:06:49,833
Really common in that space, yeah.

152
00:06:49,833 --> 00:06:53,966
And as the portable XRF becomes used

153
00:06:53,966 --> 00:06:57,433
so much more widely
customers are starting to test

154
00:06:57,433 --> 00:07:00,066
more samples that
have been fully prepared

155
00:07:00,066 --> 00:07:03,600
to get the best quality data
they can from their XRF.

156
00:07:03,600 --> 00:07:07,333
So it’s becoming really common
in sample prep facilities

157
00:07:07,333 --> 00:07:09,666
before the samples
go to the laboratory

158
00:07:09,666 --> 00:07:13,666
or samples being returned
from the laboratory,

159
00:07:13,666 --> 00:07:16,800
really good to get the
best performance,

160
00:07:16,800 --> 00:07:18,533
the best light element data

161
00:07:18,533 --> 00:07:20,266
for lithogeochemistry studies

162
00:07:20,266 --> 00:07:25,566
where the boundaries between
rock types are very strictly defined.

163
00:07:25,566 --> 00:07:28,300
Full sample prep is becoming
more and more common.

164
00:07:28,300 --> 00:07:30,500
And when you’ve invested that
kind of time in the sample,

165
00:07:30,500 --> 00:07:32,766
typically you’ll test them
in a sample cup

166
00:07:32,766 --> 00:07:34,600
or in a pressed pellet

167
00:07:34,600 --> 00:07:36,033
where you’re going to get
your best performance

168
00:07:36,033 --> 00:07:37,466
performance for light
elements again because

169
00:07:37,466 --> 00:07:38,866
if you’re shooting in a plastic bag,

170
00:07:38,866 --> 00:07:41,133
unless you are just doing
the heavier elements,

171
00:07:41,133 --> 00:07:42,866
your performance is
going to get degraded

172
00:07:42,866 --> 00:07:44,633
by the attenuation
from the plastic bag.

173
00:07:44,633 --> 00:07:47,566
Exactly. Shall we have a look
at some of this stuff?

174
00:07:47,566 --> 00:07:48,966
Yeah, let’s.

175
00:07:48,966 --> 00:07:52,133
Olympus is not a sample
preparation equipment provider.

176
00:07:52,133 --> 00:07:53,933
We specialize in what we do best

177
00:07:53,933 --> 00:07:56,300
which is manufacture portable XRF.

178
00:07:56,300 --> 00:08:00,400
Sample preparation, there’s
no one size fits all solution

179
00:08:00,400 --> 00:08:02,133
so do your research,

180
00:08:02,133 --> 00:08:03,900
have a look at what’s
on the market

181
00:08:03,900 --> 00:08:06,633
and decide what equipment you need

182
00:08:06,633 --> 00:08:09,633
to achieve your
data quality objectives.

183
00:08:09,633 --> 00:08:12,400
On the very simple side we
have things like a hammer,

184
00:08:12,400 --> 00:08:13,800
a mortar and pestle,

185
00:08:13,800 --> 00:08:15,833
run it through a sieve
to get a powder,

186
00:08:15,833 --> 00:08:19,100
put that powder in a sample cup
with a prolene film.

187
00:08:19,100 --> 00:08:23,066
More sophisticated we have this mill
which runs on AC power.

188
00:08:23,066 --> 00:08:25,433
You put your sample in the bottom,

189
00:08:25,433 --> 00:08:27,600
it takes up to a 10mm size sample,

190
00:08:27,600 --> 00:08:32,800
screw it on, press and grind.

191
00:08:32,800 --> 00:08:36,633
This has been found to be very
popular with some of our clients

192
00:08:36,633 --> 00:08:39,966
because it can also
handle very hard rock.

193
00:08:39,966 --> 00:08:42,533
This equipment here
is provided by Onscite,

194
00:08:42,533 --> 00:08:46,033
specifically designed for
the portable XRF industry.

195
00:08:46,033 --> 00:08:49,166
It’s based around an
off-the-shelf angle grinder

196
00:08:49,166 --> 00:08:53,800
with some specialist attachments
to the front end.

197
00:08:53,800 --> 00:08:55,833
This attachment is a beater mill,

198
00:08:55,833 --> 00:08:57,566
the sample goes in there.

199
00:08:57,566 --> 00:09:00,766
When it’s turned on the
beater mill crushes the rocks

200
00:09:00,766 --> 00:09:05,000
after you place the lid
on and screw it down.

201
00:09:05,000 --> 00:09:09,100
An alternative attachment
is this rotary rock grinder,

202
00:09:09,100 --> 00:09:12,566
a cutting saw which when
placed on there

203
00:09:12,566 --> 00:09:14,866
can enable you to
cut through rock faces,

204
00:09:14,866 --> 00:09:16,266
rock outcrop,

205
00:09:16,266 --> 00:09:18,800
and collect the resultant
chips and powder

206
00:09:18,800 --> 00:09:21,966
in a test-tube style attachment.

207
00:09:21,966 --> 00:09:25,100
That material can then be
passed through some sieves

208
00:09:25,100 --> 00:09:27,866
and then this torque
wrench press used

209
00:09:27,866 --> 00:09:30,200
to develop nice powdered pellets

210
00:09:30,200 --> 00:09:33,000
from which you can test
directly on the surface

211
00:09:33,000 --> 00:09:37,600
avoiding any prolene film you
might use with a sample cup.

212
00:09:37,600 --> 00:09:41,333
Another vendor are REFLEX™,
they’re part of the IMDEX group.

213
00:09:41,333 --> 00:09:45,766
They make this rock crusher
that will take 25mm feed,

214
00:09:45,766 --> 00:09:48,133
crush it down to 1-2 millimeters,

215
00:09:48,133 --> 00:09:51,333
and it will take a
kilogram of RC material

216
00:09:51,333 --> 00:09:54,200
and process it in about
a minute and a half.

217
00:09:54,200 --> 00:09:56,333
This is the REFLEX MILL™,

218
00:09:56,333 --> 00:10:01,900
feed size of 4mm down to
a fine powder in a single pass.

219
00:10:01,900 --> 00:10:07,300
Both of these systems run
on universal AC power.

220
00:10:07,300 --> 00:10:09,066
There’s also the REFLEX PRESS™

221
00:10:09,066 --> 00:10:12,400
which makes these nice pucks
that are very dense,

222
00:10:12,400 --> 00:10:15,233
very consistent and have
a nice flat surface

223
00:10:15,233 --> 00:10:17,733
so they’re quite good
for XRF analysis.

224
00:10:17,733 --> 00:10:20,933
I used the REFLEX PRESS™
on a project

225
00:10:20,933 --> 00:10:22,733
in Ireland last summer

226
00:10:22,733 --> 00:10:24,500
and the data we got on the XRF

227
00:10:24,500 --> 00:10:26,133
off the pucks was exceptional.

228
00:10:26,133 --> 00:10:27,433
And throughput?

229
00:10:27,433 --> 00:10:31,066
I know one customer doing
about 150 samples per shift

230
00:10:31,066 --> 00:10:32,800
on a 2 person operation.

231
00:10:32,800 --> 00:10:35,766
Outstanding. So you can see
that there are a wide range

232
00:10:35,766 --> 00:10:39,966
of materials and equipment
that you can use

233
00:10:39,966 --> 00:10:41,633
to do your sample preparation

234
00:10:41,633 --> 00:10:45,200
so do your homework and
see what works best for you.

235
00:10:45,200 --> 00:10:47,333
So, to conclude,

236
00:10:47,333 --> 00:10:49,766
some of you may be required
to do some sample prep

237
00:10:49,766 --> 00:10:52,833
to achieve your
data quality objectives.

238
00:10:52,833 --> 00:10:56,233
As with all previous advice,
do some test work

239
00:10:56,233 --> 00:10:58,633
to determine what level
of sample preparation

240
00:10:58,633 --> 00:11:02,566
you need for your XRF program
to be fit for purpose.

241
00:11:02,566 --> 00:11:04,533
Right, and so next time
we’ll be looking at

242
00:11:04,533 --> 00:11:07,333
QA/QC protocols to
determine fit for purpose.

243
00:11:07,333 --> 00:11:10,866
Exactly. We’re starting to
generate analytical data

244
00:11:10,866 --> 00:11:13,066
so we need to take responsibility

245
00:11:13,066 --> 00:11:14,800
for the integrity of that data

246
00:11:14,800 --> 00:11:17,100
so we’ll provide some
advice on how to do that.

247
00:11:17,100 --> 00:11:19,566
Until then, bye for now.
Bye for now.