Note: Descriptions are shown in the official language in which they were submitted.
ZO~)~4~i~
46371-~9
Canada
IDENTIFICATION OF WOOD SPECIES
The present invention relates to sampling wood to
identify wood species. More specifically the present
invention provides a method of producing an ion mobility
signature representing a wood sample and then comparing the
signature with known signatures of wood species to determine
the wood species of the sample.
When logs arrive at saw mills they are usually stacked,
and in many cases it is difficult to tell the species of a
specific log. It is known that different types of wood
species have different commercial values and thus there is
often an advantage for the saw mills to sort out different
wood species in order to maximize the values of the more
valuable woods. Today, distinguishing one species of wood
from another is often done manually either in log form or
when lumber is sorted, and this is not necessarily the most
reliable or economical method. Once the species of the wood
has been determined, there is generally a decision time of
about 2 to l0 seconds while the wood moves along a conveyor
to a position where the wood is sorted into different
species.
Z(~ ;4
Ion mobility spectrometry (IMS) is recent technolosy
which separates ionized compounds based on differences in
their drift velocity through a gas under an applied electric
fielct. This technique has the ability to produce a
characteristic spectrum of the series of high molecular
weight compounds in a matter of milliseconds. It is known
that it can produce identifiable signatures for such items as
drugs and explosives and is being developed for use by
customsr airlines and police forces to detect such
substances. Initial tests were carried out to determine if
IMS could be used to identify different wood species. A
report on these tests was published by A.H. Lawrence on
February 2, 1989 at the 75th annual meeting of the Technical
Section of the Canadian Pulp and Paper Association entitled
"Rapid Characterization of Wood Species by Ion Mobility
Spectrometry". Some tests were carried out in the positive
mode and some in the negative mode. The initial tests showed
that some wood species could be identified one from the other
provided the tests were conducted in both modes. However,
there were a number of variable parameters that did not
initially appear to be acceptable for use in the lumber
industry. First of all sampling and analyzing by an IMS
device took several seconds and this would hardly be feasible
for fast moving conveyors used in saw mills. Secondly it
seemed that only certain types of wood species could be
identified and thirdly it was not clear how such a piece of
equipment would work in a saw mill environment with saw dust,
-- 2 --
,~ .
' ' ~ .
, , :
2(~0~
other types of particles as well as vapou~s from hoth
machinery and wood are present.
Ion mobility spectrometers are known, and it is also
known that specimens analyzed by such a spectrometer can
produce different signatures, or plasmagrams as they are
sometimes referred to, which are affected by many different
variables, e.g. temperature, barometric pressure, humidity,
etc. Furthermore, when one analyzes a specimen of wood, the
wood may be dry or moist. Heartwood and sapwood from the
same wood species have different ion mobility signatures, and
there are other effects such as extraneous noises, radio
signals r vibrations etc. that may effect the signature of a
trace sample.
We have now found that by desorbing wood within a
preferred temperature range to produce trace vapours, and
ionizing the trace vapours at a further preferred temperature
range r we can measure the time of arrival of the ions and the
ion flux at a collector electrode and produce a weak electric
current signal representing an ionic signal. The measurement
can be made in the negative mode and the positive mode and
different signals are produced in the two modes. Mobility of
an ion is dependent at least partly, on the mass and shape of
the ionr as well as the charge distribution. Mobilities are
influenced by the media through which the ions travel r and by
gas density variations which in turn depend on gas
temperature and pressures. The density variations can be
., -
' ' ,''' ~ .
Z00~464
normalized by reducing the mobility to a standard temperatureand pressure and thus produce a reduced ion mobility
signature derived from the ion drift time signature for that
particular wood species. It has been found that these
different signatures can be used to identify most wood
spec:ies regardless of the fact that the temperature, and
pressure conditions vary for different locations. The
signatures are identifiable regardless of the moisture
content of the wood, and regardless of environmental
conditions. Heartwood and sapwood signatures for the same
wood species are different, but are specific for that
species.
One aim of the present invention is to be able to
analyze a wood sample within a short space of time, and
detect the wood species in less than a second. To achieve
this time and to repeat identifying different wood specimens,
more than one apparatus may be re~uired. Furthermore, under
some conditions such as analyzing a cold wood sample, then
longer times are necessary. It is a further aim to provide a
sampling arrangement that works in the environment of a saw
mill under conditions where saw dust and other dust is
blowing about under extreme noise and vibration conditions,
and still produce a signature so the wood species can be
identified.
25The present invention provides a method of producing an
ion drift time signature representing a wood species,
:: ,
.
2(~0~
compxising the steps of, heating at least a portion of a wood
samp.le at a temperature in the range of about 220 to 350~C to
desorb and produce trace vapours from the wood sample;
ionizing the trace vapours in an ionizing zone at a
temperature in the range of 220 to 350~C; pulsing ions from
the ionizing zone through a gate means into a drift region;
measuring the time of arrival of the ions and the ion flux,
for each pulse, with a collector electrode located at the end
of the drift region to produce an ionic signal, and
amplifying and averaging the ionic signal to provide an ion
drift time signature for the wood sample.
There is further provided a method of identifying a wood
specie~ comprising the steps of heating at least a portion of
a wood sample at a temperature in the range of about 220 to
lS 350~C to desorb and produce trace vapours from the wood
sample; ionizing the trace vapours in an ionizing zone at a
temperature in the range of about 220 to 350~C; determining
drift time of ions and ion flux produced in the ionizing
zone, at a collector electrode spaced from the ionizing zone,
to produce an ionic signal; amplifying and averaging the
ionic signal to provide an ion drift time signature for the
wood sample, and comparing the signature with known wood
species signatures to determine the wood species of the wood
sample.
In another embodiment there is also provided an
apparatus for producing an ion drift time representing a wood
-- 5 --
.:
.:
- '- ~ ':
z~o~
species from a wood sample comprising heating means to heat
at least a portion of the wood sample to a temperature in the
range of about 220 to 350~C to desorb and produce trace
vapours from the wood sample; gas flow means for transferring
the 1race vapours to an ionizing zone, the ionizing zone
being at a temperature in the range of about 220 to 350~C;
means to ionize the trace vapours in the ionizing zone; gate
means adjacent the ionizing zone leading to a drift region
having a collector electrode at the other side from the gate
means, the collector electrode adapted to determine drift
time of ions and ion flux in the drift region, and
amplification means and averaging means to provide an ion
drift time signature for the wood sample.
In drawings which illustrate embodiments of the
invention,
Figure 1 is a schematic diagram showing an ion
mobility spectrometer suitable for ana]yzing a wood sample
according to the present invention.
Figure 2 show six ion mobility signatures for
different wood species.
Figure 3 is a schematic diagram for air carrier
flow and purge flow for the IMS and sample line.
~!,
,
'- '
': '
2~ 9~i~
Figure 4 show six ion mobility signatures for
different samples of jack pine showing the reproducLbility of
the signatures.
Figure 5 show ion mobility signatures showing the
effects of variable moisture content for jack pine.
Figure 6 show ion mobility signatures showing the
difference between sapwood and heartwood for jack pine.
Figure 7 show ion mobility signatures at tree
loading sites under both low and high vibrations.
Figure 8 show ion mobility signatures at chipper
and canter sites with an acoustical cover on and off.
An ion mobility spectrometer (IMS) is illustrated in
Figure 1. For the purposes of experimentation a unit
manufactured by Barringer Research Limited was modified for
sampling wood. A desorber heater 10 is positioned at one end
of a spectrometer 12 and a wood sample 14 rests on top of a
filter above the desorber heater 10. A passage 16 from the
desorber heater leads through a repelling ring 18 to an
ionizing zone 20 which includes a weak radioactive source.
An electronic gate 24 separates the ionizing zone 20 from a
drift region 26. The drift region is a drift tube 28 with a
series of stacked cylindrical guard rings 30 to produce a
uniform electric field throughout the drift region 26~ A
. . ;
,
X(~0~4~,~
collector electrode 32 at the top of the drift region 26
measures the drift time of the ions and also the ion flux.
The electrode 32 produces a weak electric current which is an
ionic signal. This signal is amplified by amplifier 34 then
averaged by a signal averager 36 before being recorded on a
chart recorder 38 as a representative ion mobility signature
for the wood sample 14. A Nicolet signal averager was used
for test purposes, however, intregated averagers are used for
saw mill operations.
A sampling gas flow 40 collects trace vapours from the
wood sample 14, and transfers the vapours through a transfer
line 16, in the test machine into the ionizing zone 20. The
transfer lines 16 containing the trace vapours are heated to
prevent condensation of the trace vapour. The entire cell
is at atmospheric pressure and the ionizing source, which in
one embodiment is Ni63, a radioactive isotope, generates
certain reactant ions. These ionize a fraction of the trace
sample molecules in the sampling gas flow. As a result of a
complex interchange reaction which takes place in the
ionizing zone, the molecules of certain species of trace
vapours form ions while others do not. These ions are
prevented from entering the drift region 26 by the eleatronic
gate 24 and cannot return to the passageway 16 because of the
repelling ring 18. When the gate 24 is open, the ions
accelerate under the influence of a strong electric field
through the drift region 26 towards the collector electrode
32. The gate 24 is repetitively opened at brief intervals
,
:, : . . ~ . .:
2no~
~typically 0.2 milliseconds) emitting p~lses of mixed ions
into the drift region 26. A typical time between pulses is
20 milliseconds. As they pulse, the ions in any particular
pulse separate into their individual chemical species based
upon their differing intrinsic mobilities. The arrival of
the individual ion pulses at the collector electrode 32
produces a characteristic ion arrival time spec-trum. This
ionic signal in the form of a weak electric current from the
collector electrode 32 is amplified and then fed to the
Nicolet signal averager where it is filtered, digitized and
stacked to increase signal to noise ratio. The number of
sweeps or cycles can be varied and the average signal is
viewed on a screen in real time and subsequently displayed on
the chart recorder 38. Because each ion travels at different
velocities, the ions are separated in drift time as they
arrive at the collector electrode 32. A plot of ion
intensity as a function of drift time is referred to as a
plasmagram or signature.
A drift gas flow 42 was maintained in the drift region
26 against the ion travel direction and exited at an exhaust
44 together with the sampling gas 40. A typical time between
pulses i~ 20 milliseconds, this represents an analysis time
for one pulse of the gate 24.
With regards to heating of the sample, the desorption
temperature during tests varied from 170 to 400~C and the
tests were conducted with the negative ions analyzed and the
_ g _
.
-
'
.
~00~4~,~
positive ions analyzed. It was founcl that a temperaturerange oE a~out 220 to 350ac produced ion dri~t ~low signaturesi
that were distinguishable for different wood species. The
results indicated that in the negative mode the signatures
produced were distinctive in the desorption temperature range
of about 250 to 315~C, with a preferred temperature of 300~C
when the species were well identified. Temperatures above
315~C produced weaker peaks, and loss in distinguishing peaks.
At 350~C and above the peaks almost disappeared for some wood
species. Further tests were carried out in the positive mode
but weak signatures with peaks poorly defined were developed,
and were often common for different wood species. It was
found that in the negative mode peaks were more intense, and
plasmagrams were unique for wood species. With regards to
lS the temperatures in the ionizing zone 20, it was found that a
range of about 220 to 350~C produced satisfactory signals
which allowed one species of wood to be distinguished from
another.
.~
The velocity with which ions travel through the drift
region is given by the formula:
.,
Vd=K.E
.'
.~
'''
, -- 10 --
.
.
, :
~.
200~46~
where: Vd is the drift velocity
E is the drift field strength,
K is the scalar mobility of the ions ~cm2/V.s) for a
drift length ld which is the length of the drift
tube.
Mobility was determined from the drift time td by the
formula:
K= ~ ~td E)
secause mobility is dependent on the size of the ion,
its shape and charge distribution mobilities are influenced
by gas density variation which in turn depend on gas
temperature and pressure. These variations are normalized
out by referencing the mobility to standard temperature and
pressure.
Thus reduced ion mobility is defined as:
l~ T~ P
Ko = td~E~T~Po
-- 11 --
.
X~ fi4
':
where:
ld ~ drift tube length (Gm)
To - 273~ K
Po ~ 7~0 torr
P - drift tube pressure (usually equal to
atmospheric pressure) torr
T - drift region temperature ~K
td ~ drift time(s)
E - electric field (V/cm)
A wood sample in the form of saw dust was placed on a
filter at room temperature. The filter with the particles
thereon was then positioned over the desorber heater as shown
in Figure 1 and heated to a temperature of 300~C. The trace
vapours generated were continuously carried by the sampling
gas flow 40 into the ionizing zone 20 which was kept at a
temperature of 240~C. Pulsing occurred every 20 milliseconds
and a total of sixteen pulses were averaged for display which
corresponds to 0.32 seconds of sampling time. The appearance
of the signatures is illustrated in Figure 2. The major
peaks developed sufficiently within the 0.32 seconds so that
positive identification was carried out.
The signatures represent the average mobility of the
ions against time. The figures on top of the peaks are the
reduced mobility determined by referring the average ~obility
to standard temperature and pressure.
- 12 -
' ~' ' ~ ' '.'
2n~ ;4
In order to ensure that residual sample vapowr was
purged from the detector fast enough to facilitate the
desired sampling cycle time, a sample or carrier flow
arrangement with a purge line was prepared as shown in Figure
3. A rotating sampling head or valve 50 allows the sample
flow to pass through the transfer line 52 and enter the IMS
12. As soon as the sampling cycle is finished, the valve 50
switches through 90~, and clean air passes through the
transfer line 50 and the IMS 12 to purge all trace vapours
from the previous sample. The sampling flow flushes out
through an exit line 54.
For each detection, at least a portion of the wood
sample is heated to within the desired desorption temperature
range, The trace vapours are then carried to the ionizing
zone 20 of the IMS. A number of pulses occurs in each
detection cycle and the complete detection cycle occurs in
less than one second, preferably less than one half a second.
The IMS and transfer line purge occur after each detecting
cycle for a sufficient period of to remove all traces of the
previous vapours. A time of one second was sufficient for
the present tests. The results show that the detector can be
purged rapidly between samples thus making it acceptable for
use in a saw mill.
The steps of preheating, detecting, analyzing and
purging occurs within a time range of about 1.5 to 5 seconds.
Some of the steps, such as preheating and purging can have
- 13 -
2()~446~
some overlap, however, the preheating time is generally the
variable step as this is dependent on ambient temperature and
moisture content of the wood sample. For analyzing ~ood
samples faster than this time range, more than one IMS
detector is provided, utilizing multiple transfer lines from
one or more sampling positions.
The reproducibility of the samples were determined by
taking six separate samples of jack pine and preparing
signatures as shown in Figure 4. Some intensity variations
in the major peak and variation in the minor peaks are
observed from sample to sample. However, it has no
significant impact on identification of the species. In each
case the 1.16 Ko reading clearly stands out as being an
identifying signature.
- 14 -
'
.
~P~C
c~UI~ ~E~ aA~
Y.
t~rr~ SPlr J~ 6 (~
3~ 107
t0~ pr~c~
1C~IIto~ R~ Plr3~ t a-~
'dt~lt~ . 16~~)
~,~61 ~
b~a~ r~or ~ol~erA Lasch ~,~06~)J ~ J
~o~l
~ n~ ~ . 7 ~ ~ a ~ g ~ ~
D~13-r~ a.~ol~3 a,~,
Alplne F~ 65
lnl;~lor 9~r~ ; 8 ~1 3.
rn If~oela 1.~6~a)J 1.
l lnter~ mpl~ 9 . 28 ~ a .
~-~de~~l~ e~tol ~ e~ 3~ 2.~t
~u~ r ~ 3 ~ S . 7
~~torn 2Jea~loe~ 9 ~ 74
leoot~ am,p~
~m~b l l ~ 1 . 7 ~
Al~ln~ rl ~ ~ . O~ S
denot~ll t?~ 0~t ~roa l~r.t lor. ~l~k In t~ o~r~
2C~ 4~i4
Table 1 illustrates the wood species analyzed with the
IMS detector. The signatures for the reduced ion mobility
figures IKo) are shown for the different wood species and
also possible conflicts within the groups. The groups are
selected for ones that grow in different areas and therefore
are not likely to be mixed up in a mill. The most prominent
peaks in the signatures are the distinguishing features of
the signature.
The effects of variable moisture content are illustrated
in Figure 5. Wood samples in different states of drying were
investigated. Jack pine in three different moisture
conditions was analyzed, the green sample contains a higher
percentage of moisture than the air dried sample. However,
the signatures differ only in the time required to heat the
sample to a high enough temperature for the plasmagram to
develop. Traces B and C in Figure 5 show 10ms segments of
the signature A expanded in four consecutive analysis time
slots of 0.64 seconds. There is little difference between
the signatures for the three samples thus the moisture
content does not modify the appearance of the signature
provided the sample is heated to the required desorption
temperature preferably 300~C. Similar tests were ~arried out
with balsam fir with similar results. This means that IMS
can be used anywhere in the sequence of processing wood
products, even after the wood has been dried.
- 16 -
;'' : ' : '
' ..
: ~ :
2(~0~4~i~
Sapwood samples and heartwood samples were taken from
jack pi.ne and analyzed. The signa~ures, as shown in Figure
6, indicate that the sapwood samples contain a strong peak of
reduced mobility of 1.39 Ko~ For the heartwood samples a
1.16 Ko peak occurs and it is clear that the signatures for
heartwood and sapwood within the same species are
reproducible but differ one from the other. Other species of
wood were looked at with similar results.
In order to assess the feasibility of an IMS
installation in the field, tests were conducted at a tree
loading site and the equipment was set up in an area where
tree lengths are loaded on conveyors to be sent to a
debarker. Large amplitude shocks were experienced and
vibration signals were picked up by the IMS detector as shown
in Figure 7. Tests were also conducted at six additional
locations in a saw mill.
These locations were chosen as suitable positions in the
saw mill where the logs or lu~ber could be sorted dependent
upon wood species, and conveyed to different areas. The
locations took into account the dif~erent environmental
conditions in the mill.
Throughout the mill testing, ambient air was used
without predrying or filtering for the sample carrier flow.
The ambient air was coarsely filtered and partially dried for
the drift gas flow. No additional background peaks were
Z~0~4~.4
observed from the ambient air, as illustrated in trace 1 of
Figure 7, and no chemical interference was detected.
Acoustic and vibration effects from falling and bumping trees
were severe as can be seen in trace 2. Traces 3 and 4 are
the ~ignatures from jack pine samples from the mill run at
low and high vibration noise and trace 5 is a sample of
spruce run under low vibrational noise. In both cases
positive detection and identification is evident.
Further tests were conducted at a chipper and canter
site where many electric motors were generally runnin~
continuously. The hydraulic system was intermittent and
settled wood dust was present on all surfaces. The air was
estimated to contain about 100 particles per cubic foot.
Vibration noise was low with only occasional shocks as logs
were fed into the conveyor, however, acoustical noise levels
were severe as shown in Figure 8. Traces 1 and 2 show the
background signatures before and after an acoustical
protection cover was placed on the IMS system. Traces 3 and
4 show the signatures of jack pine and spruce on this site.
Both the jack pine and spruce samples were reliably detected.
The 2.90 Ko peak in ~igure 8 reprcsents the partially
hydrated chloride ion, ~H2O)nCl-, which is present in the
reaction region and ion mixture allowed into the drift
region. In a preferred embodiment in the negative mode,
chloride reactant ions are generated in the reaction region
from chlorinated compounds, typically ~ethylene chloride
- 18 -
"
:' '' : : '
z~o~
(dichloromethane~, introduced as a dopant into the sample
carrier gas. Under the usual operating conditions these are
partially hydrated, resulting in the reduced mobility
constant of 2.90 cm2/Vs. Upon introduction of sample
molecules into the reaction region, analyte ions are formed
at the expense of the chloride ions, and the reactant ion
concentration decreases and may become even completely
depleted.
Other reactant ions that may be used as indicators are
bromides and iodides in the negative mode, and nicotinamide
in the positive mode.
In certain locations acoustical protectors either in the
for~ of an acoustical cover or by utilizing electronic
circuits are provided to eliminate extraneous noise from
vibrations and other spurious electronic signals which are
often present in industrial locations.
The tests have shown that there is an unambiguous
signature for different wood species~ Furthermore, the IMS
application can handle wood with moisture contents varying
~0 from about 0 to 200%. The machine can operate with mill
background atmosphere and in a mill environment. For the
purposes of the test sampling was conducted with saw dust,
however, other types of sampling may be developed.
-- 19 --
2~0~i46~
Various changes may be made to the embodiments described
herei:n without departin~ from the scope of the present
invention which is limited only by the following claims.
- 20 -
: . . , .; .
''. ' ~ '' : . . ..
~ ,
