PB  200268

IMPROVED INSTRUMENTATION  FOR DETERMINATION OF  EXHAUST
GAS OXYGENATE  CONTENT

Marvin L.  Vestal,  et al

Scientific Research Instruments Corporation
Baltimore,  Maryland

March 1970
en
»
H
n
o
i
                                                 DISTRIBUTED BY:
                                                 National Technical Information Service
                                                 U. S. DEPARTMENT  OF COMMERCE
                                                 5285 Port Royal Road, Springfield Va. 22151

-------
(::
!'"'.:i
|H:i
N
Iht
ji  11
'•' I;.'V
!!;H
                                                                              !f
                                                           200   268:
                                                                * . "";*:• «*-t'•«   W.TJ
                                                                •I ''•?-;'^'4*A»  • -!4-1'
                                                                f8^<1^
                                                                 S^f'TR^?' -T" >; "6'
                                                                 S|P$^^^^'|*V'
                                                                 Jl-f.^--   - "•••••  •
           NATIONAL TECHNICAL
           INFORMATION SERVICE
              Springfield, Va. 22151

-------
                          CRC-APRAC-CAPE-ri-68-3
);; TECHNICAL REPORTS
 fifle and Suotule
 Improved Instrumentation for Determination of Exhaust Gas  Oxygenate
 Content —  SRIC
 '/. AuthorTs)
         Marvin L. Vestal; Allan G.  Day III; and William H Johnston
                                                                                  .
                                                                          3. Recipient s Catalog No
                                                                          5. Report Date
                                                                               January 1971
                                                                          6. Performing Organization Code
                                                                         8. Performing Organization Rept. No.
                                                                           SRIC 70-6
 9. Performing Organization Name and Address
   Scientific Research Instruments Corporation
   6707 Whitestone Road
   Baltimore, Maryland 21207
 12. Sponsoring Agency Name and Address                         ~
   Coordinating Research Council,  Inc.
   30 Rockefeller Plaza, New York,  N.Y, 10020; and
   Environmental  Protection Agency
.., Durham, North Carolina,	
                                                                         10. Project/Task/Work Unit No.

                                                                        TT. Contract?Gra'nFNo"
                                                                        13. Type of Report & Period Covered
                                                                         Final ~ March  1970
 15. Supplementary Notes
                                                                        14. Sponsoring Agency Code
                                                                        CRC-APRAC-CAPE-11-68
 16. Abstracts
                    This report deals with the prototype aldehyde analyzer built by Scientific Research
   Instruments Corporation  (SRIC).  With the instrument, called the AUTOSPECT, SRIC has been able
   to simultaneously analyze engine exhaust for 9 different oxygenates.  Alothough the AUTOSPECT can
   be calibrated to determine almost any desired oxygenate, the current instrument is set to determine th
   9 most prevalent aldehydes believed to be present in exhaust gas.  Work on the AUTOSPECT is aimed
   at procedural improvements and reductions to a practical portable oxygenate analyzer.  In addition
   to unburned hydrocarbons, carbon monoxide, and oxides of nitrogen, the exhaust of automotive
   vehicles contains small amounts of partially combusted materials known as oxygenates.  The
   most prevalent class of oxygenate components are known as aldehydes. Aldehydes are odoriferous,
   in some cases irritating to the eyes, and  potential sources of photochemical reactions conducive to
   the formation of smog.
 17. Key Woros and Document Analysis, (a). Descriptors

   Pollution Control
   Exhaust Gas
   Oxygenates
   odor
   smog
   photochemical
 17b. IJc-iti'lors/O,- n Cnti'id Terms

   AUTOSPECT
 7c. COSATI Field/Group
                          21E; 13B; 21D
8. Distribution Statement
Re leasable
l
to the public
19. Security Class(This Report)
UNCLASSIFIED
^.Security Class. (This Page)
UNCLASSIFIED
21. No. of Pages
56
22. Price
$3.00
     NDS-b67(1-70)

-------
                             Report No. SRIC 70-6

 IMPROVED  INSTRUMENTATION FOR
 DETERMINATION OF  EXHAUST GAS
 OXYGENATE  CONTENT

  Contract Nos. CPA 22-69-40 and CAPE 11-68

            Annual Report March 1970

       Marvin L. Vestal,  Allan G. Day III,
              and Wm. H.  Johnston

            With contributions from

C.E.  Waring, F.W. Lampe,  Gordon J. Fergusson,
J.H. Futrell, Austin Wahrhaftig, M.S.B. Munson,
Wm. F.  Biller, Edward Kratfel,  and Melvin Vanik
Prepared for the National Air Pollution Control Ad-
ministration, Durham, North Carolina, 27701,and
the Coordinating Research Council, Inc., 30 Rock-
efeller Plaza,  New York,  New York.

-------
                    TABLE OF CONTENTS


                                                          Page


TABLE OF CONTENTS                                      ii


ABSTRACT                                                 iii


INTRODUCTION                                            1


EXHAUST GAS ANALYSIS BY CHEMICAL IONIZATION
MASS SPECTROMETRY                                     3


     General                                               3

     Apolication to Engine Exhaust Gas Analyses             9


THE PROTOTYPE SYSTEM                                  13


     Chemical lonization Source                             19


     Direct Reading Mass Spectrometer                     21

     Sample Inlet System                                   23


     Calibration                                            25


RESULTS AND DISCUSSIONS                                27

APPENDIX                                                  54
                               -11-

-------
                        ABSTRACT

      Chemical ionization mass spectrometry  is shown to be a
powerful general tool for the analysis of trace oxygenates in ex-
haust emission; it provides an analytical tool for trace components
over a dynamic range  from 1 parts  per billion ( ppb) to 10,000 parts
per million ( ppm) .

      A prototype mass spectrometer employing the double focus-
ing principle of Mattauch and Herzog and featuring  fixed ion-collec-
tors at integral masses over the mass range 1  to 128 has been con-
structed and is described.   The fixed collectors make possible sel-
ection of any nine mass numbers in this range and provide full duty
cycle for each measured peak with instrumental time constant signi-
ficantly shorter than 5 seconds.

      Theoretical aoplication of the method and experimental applica-
tion of the  apparatus to the following oxygenated species in the con-
centration  range of 30-1300 ppm has been made: formaldehyde,
acetaldehyde, acrolein,  propionaldehyde,  acetone,  crotonaldehyde,
butyraldehyde, 2-butanone, n-valeraldehyde, 2-pentanone,  3-pentanone,
ethyl-butyraldehyde, and benzaldehyde.  It is shown that in all cases,
analysis is feasible via the M + 1 ion ( protonated parent compound)
using reagent ion CH5  except that,  with this reagent,  interferences
occur for acrolein and crotonaldehyde necessitating the use of a
different reagent for these species.
                              -111-

-------
      Calibration of the mass spectral apparatus for all the above
compounds has been made.  Sensitivity of response for the various
compounds range from 5-108 millivolts per ppm of oxygenated com-
pound.   With the exception of the larger aldehydes this response is
practically all due  to the M + l ion.   In the case of the larger aldehydes
loss of H2O from the protonated molecule,  forming the M-17 peak,
is the dominant  process.

      Preliminary measurements were sufficiently encouraging that
the breadboard stage was skipped and a prototype  research unit was
designed and built.   Further work requires the redesign  of one por-
tion of electronics,  optimization of operational parameters,  tests
in cooperation with Scott Laboratories  on exhaust  from a  current
model automobile under cyclic conditions,  tests for other compon-
ents such as  NOX,  exploration of negative ion  reactions and design
of a simplified unit for optimum  utilization by untrained personnel
and at minimum cost.
                              - iv-

-------
INTRODUCTION


     The goal of this research is the  development of an instrumental

method of analysis for oxygenated hydrocarbons in  engine exhaust.
This method is based on the combination of the chemical ionization

technique developed by Field and coworkers combined with the direct
reading mass spectrograph developed at Scientific Research Instru-
ments Corporation  ' . A bibliography of publications  on chemical

ionization mass spectroscopy is given in the Appendix.
     The theory of analysis by chemical ionization is presented in

Chapter II of this report.  This theory shows the enormous enhance-

ment in sensitivity and specificity which is possible with chemical
ionization as compared to electron impact ionization.


     The prototype  system combining chemical ionization with the direct

reading mass spectrograph for  the rapid,  continuous analysis of oxy-
genated hydrocarbons in engine exhaust is described in Chapter III.

This prototype system employs a chemical ionization source which

may be operated at source pressures  as high as 10 torr.  Mass  analysis
is performed using a. small Mattauch-Herzog type double-focusing mass
spectrometer.  Fixed collectors are provided for each mass in  the range
from 1 to 128 amu.  Any  nine of these masses may be  monitored contin-
uously using nine independent electrometers and  readout channels.
( 1)  This method is covered by patent rights of Esso Research and Engi-
     neering Company with exclusive license to Scientific Research In-
     struments Corporation, Baltimore,  Maryland.
                             -1 -

-------
    Chemical ionization measurements have been conducted on a

series of the aldehydes and ketones  identified by Oberdorfer    as
the more abundant oxygenated species occurring in engine exhaust.

The compounds studied include formaldehyde,  acetaldehyde,  pro-

pionaldehyde,  n-butyraldehyde,  iso-butyraldehyde,  n-valeraldehyde,

iso-valeraldehyde,  ethyl butyraldehyde,  acrolein, crotonaldehyde,

benzaldehyde, acetone, butanone, 2-pentanone, and 3-pentanone.

These experimental results show that the technique  should be cap-

able of detecting most of the oxygenated  hydrocarbons at levels from

300 parts per million down to 10 parts per billion in engine exhaust using

methane as a chemical ionization reagent.  With the exceptions of acro-

lein and crotonaldehyde no interference from other compounds in engine

exhaust is indicated by our data taken in conjunction with the results of

other  investigations described in the literature.  In the case of acrolein

the crotonaldehyde  some interference from hydrocarbons is encountered

when  methane is used  as the reagent.
(2)  P.E. Oberdorfer,  "The Determination of Aldehydes in Automobile
     Exhaust Gas, " Paper 670123 presented at the SAE Automotive
     Engineering Congress, Detroit,  Michigan, January,  1967.
                             -2-

-------
    Chemical ionization measurements have been conducted on a

series of the aldehydes and ketones  identified by Oberdorfer    as

the more abundant oxygenated species occurring in engine exhaust.

The compounds studied include formaldehyde,  acetaldehyde, pro-

pionaldehyde,  n-butyraldehyde,  iso-butyraldehyde,  n-valeraldehyde,

iso-valeraldehyde,  ethyl butyraldehyde,  acrolein, crotonaldehyde,

benzaldehyde, acetone, butanone, 2-pentanone, and 3-pentanone.

These experimental results show that the technique  should be cap-

able of detecting most of the oxygenated hydrocarbons at levels from

300 parts per million down to 10 parts per billion in engine exhaust using

methane as a chemical ionization reagent.  With the exceptions of acro-

lein and crotonaldehyde no interference from other compounds in engine
exhaust is indicated by our data taken in conjunction with the results of

other  investigations described in the literature.  In  the case of acrolein

the crotonaldehyde  some interference from hydrocarbons is encountered

when  methane is used  as the reagent.
(2)  P.E. Oberdorfer,  "The Determination of Aldehydes in Automobile
     Exhaust Gas, " Paper 670123 presented at the SAE Automotive
     Engineering Congress, Detroit,  Michigan, January,  1967.
                             -2-

-------
EXHAUST GAS ANALYSIS BY CHEMICAL IONI/ATION MASS
SPECTROMETRY
General


     Chemical ionization mass spectrometry produces ionic derivi-

tives of the compound whose analysis is required by a sequence of

reactions involving:


1.  Electron impact ionization of a reagent gas.


2.  Ion-molecule reactions of the reagent gas to produce a set of

reagent ions that are stable with respect to further reaction with the

reagent gas .


3.  Reaction of the reagent ions with the compound whose analysis is

required to produce ions characteristic of this compound.


By  contrast, conventional  mass spectrometric analysis involves direct

ionization of the unknown by electron impact  at pressures sufficiently

low that  collisions  of the resulting ions with molecules do not occur.

Thus if we denote the  reagent gas by M,  the stable reagent  ions by

R  , and  the substance to be analyzed by X, we have the following

ionization mechanisms:
A.  Chemical Ionization   M + e  —> M^   +  2e                    ( l)


                          Mi+ + M  —> R+ + Ni                  (2)


                          R++ X —> X+ + R                    (3)
                                -3-

-------
B.  Direct lonization    X + e —> X: +  2e                         (4)

In these ionization mechanisms, Mj   represents the set of reactive
ions of reagent gas M, N- is a set of neutral products that need not
concern us further,  R   is a reagent ion (or set of reagent ions) ,  X.
is the  set of  ions characteristic  of the  unknown and the reagent gas,
and X.  is the set of ions characteristic of electron impact ioniza-
tion of the unknown X.
      Note that in the case of chemical ionization the set of ions
X^  is produced without the simultaneous ejection of an electron that
occurs in the direct ionization of X to form the set of ions X- .
This  is a very  important difference in the two types of ionization
since the absence of a free electron in ( 3 ) imparts much more  severe
energy restrictions to this  reaction than is the case for (4 ).   The
net result is that one expects the set of ions Xj  to be a much simpler
set than  X-  ; i n fact, one can envision systems in which X^  is  a single
ion, a situation practically never obtaining for the  set X;  produced
by (4).   These expectations  have been  confirmed by many experi-
ments utilizing a number of reagent ions, so that there is no doubt
that chemical ionization mass spectrometry provides the analyst with
a much simpler mass spectrum  to interpret.   This greater simplicity,
in turn,  leads to more  reliable identification of the unknown compounds
present in any  system to be analyzed.
                                  -4-

-------
      Sensitivity


      In addition to simpler mass spectra, chemical ionization mass

spectrometry is a considerably more sensitive method of analysis for

compounds present at the part-per-million (ppm) level in a gaseous

mixture.   Consider, for example,  an unknown X present at a concen-

tration level of 1 ppm in some carrier gas.  If we were to analyze for

X with conventional electron impact mass spectrometry we would en-

vision the following conditions:


      Total ion-source pressure, P-y  < 10 5 torr .
      Concentration of sample in ion source, n-p  = 3 x 10l
      (molecules/cm3, for P  in torr) .
     Electron currents Ie < 1 ma =  10~3 amperes.
     Electron impact ionization cross-section »|fl{,10  15 cm2.
     ( nominal value for 70 ev electrons)
     Electron path in ion source, LiefSi 1 cm.
     Detection efficiency, EsJl% (efficiency with which ions  produced in
     the source are transmitted through the mass analyzer and detected
     at the ion collector) .
Under these conditions the total ion current detected is given by


      IT= <^i nT IeLeE                                                (5)


        = 10"15( cm2) •  3 x 1011  ( cm"3)  •  10"3 ( amperes) • 1 ( cm) . 10"2


        = 3 x 10  9 amperes


                                   -5-

-------
The total ion current for the unknown X present at a concentration



level of  1 part per million is  given by
           I  -  10"6 IT = 3 x 10"15 amperes                         (6)
            3v        -L





This current is  normally distributed among a number of fragment ions



produced by electron impact ionization.







     The total ionization efficiency in the  electron impact ionization



can, in principle, be increased by increasing the electron current, the



ion source pressure,  or the detection  efficiency.  However, the max-



imum ion current which can be handled in  most mass  analyzers without



introducing non-linearity due to space charge effects  is about 3 x 10 9



amperes.   Therefore, little additional enhancement in sensitivity can be



obtained by increasing the total ionization  efficiency.






     In chemical ionization mass spectrometry the total pressure in the



ion source is  typically about 1 torr.   The  intensity of the primary ioniza-



tion to produce the ions M^  according to reaction ( 1) is still given by



Equation ( 5) .   As can be seen from Equation ( 5) above, much smaller



values of either electron current or detection efficiency can be employed



while still realizing the maximum total ion beam that  can be  utilized in



the mass analyzer.   As a result,  the  small orifices required in the



chemical ionization source to obtain high pressures do not limit the  sen-



sitivity of the technique since more than adequate electron beams and ion



beams can be transmitted through the necessary orifices.
                                    -6-

-------
      In chemical ionization the reagent ion intensity is given by


                                   i)                                   (7)


where Im  is given by Equation ( 5) ,  n-p is the concentration of reagent gas

in the  ion source, 0"^ is the cross-section for the ion-molecule reaction

    > 10  15 cm2), Lj is the average  ion travel distance  in the ion source

    sfcl cm) .   The exponent in  equation (7) may be expressed by
             = P2nT Li = 3 x 10<> PT L.
                                                                        (8)
             > 3 x 1016 ( 10"15) (1) PT = 30 P
where Pnn is in torr.   Thus the ions initially produced by electron impact
(M^ )  are quite completely converted  to reagent ions ( R ) if the ion source

pressure is a few tenths of a torr or more, and l( R  ) f& Irp .


      If X is at a sufficiently low concentration that its ionization by re-
action ( 3) does not significantly deplete the concentration of reagent ions,

it is easily shown that the ion  current  produced from the species X is

given by
                                                                        (9)
Where n-^ is the ion source concentration of X ( molecules /cm3) , ff"3 is
the cross -section for  reaction ( 3) , and Lj is the ion path length.


     At low sample concentrations , ff"3 n,,.L^ « 1 ,  and equation ( 9)  reduces
to
                                   -7-

-------
  Ix = IT .
                                                                    10)
Experimentally,  the value oftf^ Lj. _> 10 15 cm3 for many of the ion

molecule  reactions involved.  For an ion source pressure  Pry,, and

a concentration of X of 1 ppm, Equation ( 10) becomes:



             Ix = IT  .  3 xlO16 PT  . 10"6 (molecules/cm3). 10"15 (cm3) ( 11)


                = 3 x 10~5 PT .  IT
      Comparing with Equation  6  we see that
Iv (CHEMICAL IONIZATION)
 A
Ix (ELECTRON IMPACT)
                                         •? ™
for the same total ion current,  at the maximum operating pressure of the

prototype chemical ionization system ( 10 torr)  chemical ionization is

at least 300 times as sensitive as electron impactionization in terms of

the ion current produced from a trace component for a given  total ion

current.


     Linearity



     The ion current produced from reactive species X in the reagent

gas is given by Equation (9).   At low concentrations the ion  signal is

directly proportional to  concentrations of X as given in equation ( 10) .

For Equation ( 10) to be valid within 5%,  it is  required that
                                  -8-

-------
                (J^ nx L,i <_ 0.05



                                                                   (13)



                    nX <  5 x 1013 molecules/cm3






at higher concentrations a significant fraction of the reagent ions are



destroyed by reaction,  and the exponential equation ( 9)  applies  rather



than the linear equation ( 10) .  The implications of this analysis of



sensitivity and linearity is summarized in Table I.







Application to Engine Exhaust Gas Analyses







     The specific problem to be solved in these analyses is to measure



oxygenated hydrocarbons at low concentrations in  a complex mixture of



gases, the principal component of which is nitrogen.  Nitrogen,  the



natural carrier gas of this complex mixture, is not itself a suitable  re-



agent gas for the analysis of oxygenated hydrocarbons.  A suitable re-



agent gas must, therefore, be added to the mixture.   At first sight this



might seem to  involve a serious dilution of the sample.   However, the



fact that  molecular nitrogen ions will undergo charge transfer reactions



with practically any reagent gas we would envision easily circumvents
this difficulty.  For example, suppose methane, having reagent ions



CH5  and C2H5 , had been added to the mixture as the reagent gas.



Any electron impact  ionization of the nitrogen will be followed rapidly by
               N2 + + CH4  	> CH4+ + N2                           (14)
and
               N2  + CH4  —-> CH3 + + N2 + H                        (15)
                                   -9-

-------
 TABLE I.  SENSITIVITY AND LINEARITY OF CHEMICAL IONIZATION
            ANALYSES AS A FUNCTION OF TOTAL ION SOURCE PRESSURE
 Pressure       Minimum Detectable Concen-   Maximum Allowable Concen-
    (torr)        tration ( parts per billion) ( 1)   tration ( parts per million)( 2)
    0.2

    0.5

    1.

    2.

    5.

   10.

   20.

   50.

  100

  200

  500
1,000
 ( 1)  Corresponding to 3 x 10  15 amperes mass analyzed ion current for
      trace species and  3 x 10  9  total mass analyzed ion current.
 (2)  For deviation from linearity <  5%.
150
60
30
15
6
3
1.5
0.6
0.3
0.15
0.06
0.03
7, 500
3,000
1, 500
750
300
150
75
30
15
7.5
3
1.5
                                  -10-

-------
thus producing the same principal active ions of methane that will

follow electron impact ionization of methane.   These ions  will not

react with molecular nitrogen but will react rapidly with methane  to

produce the normal reagent ions CH5  and C2H5 .   Therefore,  with

the exception of production of a slightly different ratio of reagent ions,

ionization  of the nitrogen is equivalent to  ionization of reagent gas and

there is no loss of sensitivity.



      The basic nature of the oxygenated compounds to be analyzed

suggests that a particularly appropriate chemical ionization reaction

is that of proton transfer from a suitable reagent ion to the oxygenated

hydrocarbon, viz:
                 XH+ +     C = 0  — >C-OH  + X                (16)
This reaction will occur readily for any hydrogen containing reagent

ion provided that
>
                                    P(X)
where P (X) is the proton affinity of compound X.  Of course,  if the pro-

ton affinity difference is too great the exoergicity of the reaction may
                                              TD ^      I
result in considerable fragmentation of the ion   .^C-OH  .   Some frag-

mentation may be desirable for clarity  of identification when mass  inter-

ference occurs, but, in general,  extensive fragmentation is to  be avoided.

The choice of reagent gas  is thus directed towards one having reagent ions
   i
XH  such that P(X)< P(  . C = O) but withAP being as small as possible.
                                 -11-

-------
      Ideally,  the reagent gas should produce reagent ions that
would not protonate other compounds in the gas mixture, such as
olefins, aromatics,  paraffins, carbon monoxide,  etc.   Although
some selectivity in this regard is possible, complete selectivity of
protonation of oxygenated hydrocarbons will not,  in general,  be
possible.   Hence, we must choose a reagent gas  whose reagent ions
and product ion produced by reaction with other species present do
not by mass interference obscure the protonated forms  of the oxygenated
hydrocarbons .

      These considerations lead to the conclusion  that methane,  with
reagent ions CH5  and C2H5  (M/e 17 and 29, respectively) ,  is a reason-
ably good initial choice for a reagent gas for  the oxygenated hydrocarbons.
The pertinent  chemical ionization reactions are the following:
              CH5  +    C = 0 — ^C-OH  + CH4                   (3A)

                                         + C2H4                   (3B)

                                          + + C2H6                  (3C)
where,  depending on the specific energetics, dissociation of the ionic
products may occur to some extent.   Chemical ionization spectra for
methane reagent reacting with a number of aldehydes and ketones are
presented in Chapter IV.

-------
THE PROTOTYPE SYSTEM








     The prototype system for analysis of oxygenated hydrocarbons  in



engine exhaust gases consists of a chemical ionization source coupled with



a direct  reading mass spectrometer.  A heated gas inlet system is pro-



vided for interfacing between the exhaust gas sampling point and the  inlet



of the mass spectrometer.   A schematic diagram is given in Figure 1 of



the ion optical system for the  combination  of the  chemical ionization source



with the  direct reading mass  spectrometer.








     The mass analyzer employs the Mattauch-Herzog double focusing



geometry with a 6" radius electrostatic analyzer and a fixed collector for



each mass from 1 to 128 amu.   This system was designed to be operated



normally in the direct reading mode in which a selected  set of masses are



monitored continuously.  However,  provisions for electrically scanning to



mass 250 are included in the  design and scanning has been used during the



initial exploratory research.








     The chemical ionization  source and differential pumping systems are



designed to operate at source  pressures as high as 10 torr for efficient



chemical ionization of the oxygenated hydrocarbons of interest.   With these



high ion  source pressures,  the pumping system maintains the pressure  in



the ion source housing at a  suitably low pressure (typically 10 4 torr) to



avoid electrical breakdown.    The analyzer section is isolated from  the



source section by a differential pumping slit and  maintained at sufficiently



low pressure  (typically < 10   torr) to avoid degradation of performance by



ion-molecule  collisions in the analyzer.   Sufficient testing has  been accom-



plished to assure that  the differential pumping requirements are met by the
                                   -13-

-------
CD
                                                                    v
                                                                    N
                                                                    rt

                                                                    «J
                                                                    nt
                                                                    BO
                                                                    3
                                                                    rt
                                                                    C
                                                                    o
                                                                   -OJ
                                                                    M
                                                                    c
                                                                    o
                                                                    rt
                                                                    o



                                                                    1)
                                                                   a
                                                                   o
                                                                   c
                                                                   o
                                                                   n)

                                                                   oo
                                                                   (0

                                                                   Q

                                                                   u

                                                                   •4-*
                                                                   (4


                                                                   V

                                                                   U
                                                                   w:
                                                                   t,
                                                                   3
                                                                   &>o
      -14-

-------
present design.  Heaters are provided for the chemical ionization
source and the glass inlet lines to allow operation at temperatures from
100°C to 350°C to minimize the anticipated problems in accurately  sam-
pling the raw engine exhaust.

      In Figure 2 a pictorial sketch is given showing the relative locations
of the principle components  of the system.   The ion source housing and
electrostatic analyzer housing are separately pumped by high speed
4 inch diffusion pumping  systems located as  shown in Figure 2.   The
mechanical pumps backing the diffusion pumps are located in the bottom
of the cart supporting the mass spectrometer.  This cart is mounted on
heavy duty casters and contains the complete mass spectrometer inlet
system and pumping systems.  The electronics for the system are loca-
ted in a separate caster-mounted  cabinet.  The complete system is  designed
to be readily mobile.

      A photograph of the principal components of the chemical ionization
direct reading mass spectrometer is shown in Figure 3.   These compo-
nents are contained  within the vacuum housing shown schematically in
Figure 2 in the relative locations within the system shown in Figure 1.

      A photograph of the completed prototype system is shown in Figure 4.
The  electronics cabinet on the left in Figure  4 contains all of the readout
and control systems for the  prototype unit.   The  cart on the right contains
the direct reading mass spectrometer with chemical ionization source and
the associated vacuum  systems and sample inlet  system.
                                 -15-

-------
                                          V
                                          .c
                                          o

                                          on
                                          44

                                          4)


                                          i


                                          §
                                          U
                                          fH

                                          a-

                                          y
                                          c
                                          V
                                         .c
                                          to
                                          c
                                          o
                                          o

                                          O
                                         r-i  0>

                                          V  N
                                          Ml rt
                                          c  c
                                         •s
                                         **  -
                                         «  2
                                         •* -
                                         •C .2
                                         5 ^
                                         O  N

                                         s§
                                         1X1  rt

                                         «  2

                                         S  6
                                         Bp  «)

                                         tt  u
-16-

-------
                                                                    0)
                                                                   TJ  W
                                                                    ti  to
                                                                    nJ  nt
                                                                    0)

                                                                    o
                                                                    4)
                                                                    fl)
                                                                       r-i

                                                                       o
                                                                         .

                                                                   -•?
                                                                    „  s
                                                                    K  «>
                                                                    3  -C
                                                                    Jo  U
                                                                       I,
                                                                    to  O  «
                                                                    -M  -M  rr
                                                                    C  u  ^
                                                                    D  0)  H

                                                                    g  -  8
                                                                    n  °  ^
                                                                    O.  ,  *j
                                                                   ro

                                                                    0>
                                                                    GO
-17-

-------
Figure 4.  A Photograph of the Completed Prototype System.

The electronics cabinet on the left contains all of the readout and
control systems and the cart on the right contains the  direct reading
mass spectrometer with chemical ionization source and associate
vacuum systems and sample inlet systems.
                               -18-

-------
Chemical lonization Source








      The chemical ionization source was designed to allow both



chemical ionization and electron impact  ionization with the prototype



system.   A detailed diagram of the chemical ionization source is



shown in Figure 5.   This source design is very similar to the con-



ventional Nier type ionization source for a magnetic mass spectro-



meter with several detailed modifications to allow its  use for chemical



ionization.








      The sample flows into the source through a hole  in the repeller.



The  repeller and electron trap are sealed to the source block by cera-



mic  washers to minimize extraneous leaks of gas  from the source.



The  only substantial gas flows out of the source are through the orifices



provided for the entrance of the electron beam and the exit of the ion



beam.  The electron beam entrance is a 0.020" diameter hole and the



ionization exit is a slit 0.002" x 0. 125".   The electron beam is  mag-



netically collimated using a 200 Gauss source magnet  built up from



Alnico V bar magnets and soft ion pole pieces .  The electron emitting



filter is a Rhenium ribbon 0.001" thick by  0.030"  wide welded to 0.040"



diameter molybdenum.   The other metal parts in the  ionization source



are fabricated from Inconel 600 alloy and the ceramic insulators are



high alumina ceramic.








      Heaters for the source and sample  inlet lines are provided and



the system is bakeable to 350°C.   The final slit in the ionization gun



is  sealed to a baffle  separating the source  and the electrostatic analyzer



housings .   This slit  is 0.005" x 0.125" and  constitutes the only sub-



stantial gas flow orifice between the  two  housings.  These two housings



are separately pumped by 4" oil diffusion pumps backed  by 140 litre
                                -19-

-------
                            SAMPLE
   ION SOURCE BLOCK
                MOUNTING PLATE
MAGNET POLE
     DIFFERENTIAL
      PUMPING SLIT
ION BEAM
                        MAGNET POLE
                                               ELECTRON ENTRANCE


                                               ELECTRON BEAM
     BAFFLE BETWEEN SOURCE AND ELECTROSTATIC ANALYZER HOUSINGS
           5.
     Schematic  Diagram Of  The  Chemical  lonizarion

     Source For  The  Prototype System  For  Analysis Of

     Oxygenated  Hydrocarbons  In  Eng ine Exhaust.
                             -20-

-------
per minute mechanical pumps.   This differential pumping system
allows operation with source pressures as high as 10 torr while
maintaining the source housing at pressure below 10 3 torr and
the electrostatic analyzer housing below 10  torr.   Total ion
beams in excess of 10 9 amps have  been obtained with this ioniza-
tion source.

Direct Reading Mass Spectrometer

      The mass spectrometer is based on the original Mattauch-
Herzog design of a double focusing mass spectrometer.   The electro-
static analyzer is a 6" radius cylindrical  condenser providing a de-
flection angle of 31.8°.   This assembly is  constructed  of
Inconel 600 and high alumina ceramic insulators and is bakeable
to 350°C.  An Alnico V permanent  magnet  provides a uniform mag-
netic field of 3500 Gauss across a half-inch gap.   The Mattauch-
Herzog geometry was chosen because it is  double focusing for all
masses simultaneously.    The  foci for individual masses fall in  a
precisely determinable plane and the location along the plane is prop-
ortional to the square root of the mass.   This  view allows the precise
a priori determination of the relative locations for collectors for spe-
cific  individual masses.

      The principal unique feature of the mass analyzer  is the multiple
collector  array used as  a  substitute for the  traditional photo-plate
detector in the Mattauch-Herzog geometry.  This  collector  assembly
consists of separate detectors for each  unit mass in the range from
mass 1  to mass 128 amu.   A close-up photograph  of the detector
assembly is shown in Figure 6.  An electrical connection from each
collector  is brought out  through a vacuum feedthrough to a connector
                                -21-

-------

-------
panel.  This connector panel, which is similar to a miniature



telephone switchboard, provides the means  for selecting the masses



which are monitored.   Ten output channels are provided each con-



sisting of a solid state electrometer mounted on the switchboard panel



and a meter readout located on the front panel of the electronics cab-



inet as shown in Figure 3.   The full scale sensitivity of these channels



is continuously adjustable over a  1000:1  range from the front panel.



Individual mass collectors  selected for monitoring are connected to



the input of one of the electrometers through a short jumper wire.



The unused collectors are grounded by the use  of special grounding



clips  provided.   By merely rearranging the jumpers on this  switch-



board the set of masses monitored can be changed to correspond to a



particular measurement.








Sample Inlet System








      The sample inlet system for chemical ionization measurements on



exhaust gases is shown schematically  in Figure 7.   This system is



designed both for monitoring  of actual exhaust gas samples directly



from vehicle exhaust and for  calibration of  the  system  using artificial



exhaust gases.   The entire system through which exhaust gases pass is



surrounded by a heating jacket which will maintain the  system at approx-



imately 150°C.   An infusion  pump and syringe is  provided for controlled



injections of calibration samples  into the gas flow.   The gas flow  rate



is monitored by a mass flow meter.    A  glass manifold is provided for



coupling  between the main flow of exhaust gas and the chemical ioniza-



tion mass spectrometer.   The glass manifold is designed to  operate  at



a nominal absolute pressure of 1/2  atmosphere.   A  small  sample of



the exhaust gas  stream is drawn off  into this manifold through a fixed
                                  -23-

-------
                                             MASS
                                             FLOW
                                            METER
               VENT OR
                 PUMP
                                                             I



MUST GAS-—*-
I

LJCATlMf* 1 A f* L* CT
nt-A 1 Irib JAOK tT
	 1 	










1

                                                                  I
                                       FIXED LEAK
               • SYRINGE


                INFUSION PUMP
           CHEMICAL
           IONIZATION
           MASS
           SPECTROMETER
                                                             /
                                                               /
PRESSURE 6AUOE- 0-760 MM.

SUB-ATMOSPHERIC PRESSURE REQ.
                                                       REAGENT GAS
                                        VARIABLE CONDUCTANCE VALVE
                      MECHANICAL VACUUM PUMP
      7
Schematic  Diagram  Of  The Sample  Inlet And  Calibration
System   For The  Prototype System  For Analysis  Of
Oxygenated   Hydrocarbons  In   Engine  Exhaust-
                               -24-

-------
leak.  A second fixed leak connects the glass manifold to the
chemical ionization mass  spectrometer.   The reagent gas is
metered into the glass manifold through a sub-atmospheric pres-
sure gas regulator.  The total pressure in the glass manifold may
be adjusted by using a variable conductance valve between the  mani-
fold and  a mechanical vacuum pump.   This valve serves as a  fine
control on the pressure  in the manifold and hence the pressure in
the chemical ionization mass spectrometer.

Calibration

     The inlet system was designed to allow rapid calibration of the
entire  prototype  system.  A sample of artificial exhaust gas or inert
gas is  passed through the main exhaust gas line at a high flow  rate,
typically 15 litres per minute.   A sample  of aldehyde or ketone, or
other compound for which a calibration measurement is required,  is
placed in the syringe and fixed in position on the infusion pump.  With
a known  infusion rate and a known flow rate of the exhaust gas, precise
mixtures at the ppm level can be obtained.  A sample of this mixture
is drawn into the glass manifold where the reagent gas is added in a
proportion which can be read from the absolute  pressure gauge pro-
vided and the resulting mixture analyzed in the chemical ionization
mass spectrometer.  The  entire inlet system is operated hot  (typically
150°C) during  calibration measurements in an attempt to avoid errors
due to  condensation or polymerization on the walls.
                                -25-

-------
      This  same inlet system is also used in the actual exhaust
measurements.  With the syringe and infusion pump in place, the
calibration for a particular compound can be checked at any time
by injecting that compound at a known rate into the exhaust gas
stream.
                                 -26-

-------
RESULTS AND DISCUSSIONS








      Chemical ionization mass spectral measurements have been



carried out on a number of aldehydes and ketones.   In most of this



work methane was used as the reagent  gas but experiments  have also



been carried out using propane,  iso-butane, neo-pentane, and water



as reagents.  The mass spectra obtained in a series of measurements



using methane are given in Figures 8 through 25.   The relative sensi-



tivities observed in these measurements  are summarized in Table II.








      These measurements were carried  out using a cold inlet system



and some of the variations in the sensitivities observed are probably



due to difficulties  with the inlet system.   Also  some of the  spectra



show residual contamination from compounds run in the previous



measurements.   When high  concentrations of the sample were used,




for example, Run  No. 2 on crotonaldehyde, Run No. 1 on benzaldehyde



and Run No. 13 on acetaldehyde, some reduction in the  sensitivity is



observed as a result  of using up a significant amount of the  methane



reactant ions.  Below concentrations  of about 300 parts per million



this effect  is not observed and it is expected that the signal  should be



linear with concentration at lower concentrations.    With the exception



of Run No.  17 this series of  measurements was  conducted at a  source



pressure of 1 torr, source temperature of 150°C,  and electron em-



ission of 400 microamps.








      The aldehyde-methane  mixtures  were prepared on a stainless steel



vacuum line.  A  sample of liquid reagent grade aldehyde was expanded



into the evacuated line and the pressure measured with a mechanical



manometer.  A small sample of the vapor (0.5 cc)  was isolated and



then further expanded into an evacuated 500 cc glass bulb.   The glass
                                  -27-

-------
TABLE II: RELATIVE SENSITIVITIES MEASURED IN A SERIES
OF CHEMICAL IONIZATION MEASUREMENTS  ON ALDEHYDES
AND KETONES USING METHANE REAGENT.
                                                   Sensitivity
Run
No.
13
15
6
14
12
2
7
5
11
10
9
8
3
4
1
Compound
Acetaldehyde
Acrolein
Propionaldehyde
Acetone
Crotonaldehyde
Crotonaldehyde
n-Butyraldehyde
iso-Butyraldehyde
2-Butanone
n-Valer aldehyde
iso-Valeraldehyde
2-Pentanone
3-Pentanone
Ethyl butyraldehyde
Benzaldehyde
Molecular
Weight
44
56
58
58
70
70
72
72
72
86
86
86
86
100
106
Concentration
(ppm)
350
140
270
270
80
40,000
200
35
55
50
40
60
100
30
1300
mv / p pm
M + l
24
69
27
36
29
0'.3
6
80
33
5
2
20
10
8
4.5
)
Total
24.4
70
28
37
31
0.4
19
108
34
19
15
21
11
33
5
                           -28-

-------
bulb was then filled with methane to 1 atmosphere of pressure.  By


this technique the concentration of the aldehyde  in the methane is


given by
                          V2  • P2   1000  •  14.7





Where Vj is the volume before expansion, V2 is the total volume, in-


cluding the 500 cc bulb, Pj is  the measured pressure of the aldehyde


vapor in ps ia and P2 is the methane pressure.   In using this static


sample preparation it is possible that the actual final concentration of


aldehyde may be somewhat lower than calculated if sufficient time for


equilibration is  not allowed.   However,  the concentration should not be


higher than calculated.






     The measurement on formaldehyde used Baker analyzed reagent


grade formalin solution containing 37.8% HCHO stabilized with  13. 1%


methanol.   The methane chemical ionization spectrum for this mix-


ture is given in  Figure 8.   The concentration of formaldehyde in this


sample was approximately 1000 ppm.  The protonated  formaldehyde


ion is at mass 31 and the protonated methanol ion is at  mass 33.


The only interference with the protonated formaldehyde that  is ex-


pected to occur  is the double C13 isotope  peak from C2H5 ; this may


limit the detection limit for formaldehyde to about 0. 1  ppm.





     The chemical ionization  spectrum for  1300 ppm benzaldehyde  in


methane reagent is given in Figure  9.   The principal benzaldehyde


ion is P+1 at mass  107.  Impurity peaks from the earlier run on formalin
                              i

solution are visible at masses 31  and 33.   It may be noted that the  mass



17 peak from methane is rather small compared to the  other peaks  in
                                 -29-

-------
   XI
Sensitivity
   X10
                     •H—I—h
-I—h—1—h
                          ±
                          V
                                         C
                                        A-LLLi
                                                 2
           ttk
                                                           \
                                          H—I—I—I—I	1—I—I—I—\-
                                       t
                 „ J M
                     H	1	1	1	H—I	1	1	1	(-
              H	1	1	1	h
  H	1	h
                              CLEVITE CC
                              H—I—I—f
H	1	1-
                               20
     30
                                                          40
                                     MASS NUMBER
                                   (Atomic Mass Units)
                    Figure 8.  Chemical lonization Mass Spectrum for
                               Formalin Solution with Methane Reagent
                               at 1 torr and 150°C.
                                  -30-

-------
                                   rt
                                      w
                                       u
                                             0)
                                             G
                                             a)
                                             0)
                                             T)
                                             >N
                                             J3
                                             V
                                             T3
                                             •—i
                                             rt
                                             N  •
                                             C -«


                                             « £

                                             ^ -5
                                             o 8
                                             M-l Pt
 3 C!
 ^ 3

tJ Pi
 (U

 an"
W U
 w in
 rt •->
2*

 gs
^  M
 rt  n
 N  O
•r-l -U

 g-
I-H ^
^H  RJ
 rt ^
.a  «
 3  «
 S  M
 0)  nl
J3  
                                              3
                                              Ml
            o
            rj

            X
 a>
W
          -31-

-------
the spectrum,  indicating that a substantial fraction of the reagent
ions have been used up by reaction at this concentration of benzaldehyde.

      The chemical ionization spectrum for 4% crotonaldehyde in methane
reagent is given in Figure 10.  The major crotonaldehyde peaks in this
result are the P+1 at mass 71,  and  the P-l at mass 69.   A significant
amount of the protonated dimer ion  is observed at mass 141.  At this
relatively high sample  concentration, it may be noted that the methane
reagent ions have virtually disappeared.

      The chemical ionization spectrum for 100 ppm of 3-pentanone in
methane reagent is given in Figure  11.   The only significant peak pro-
duced from the 3-pentanone is the P+l ion at mass 87.  Impurity peaks
from the early runs of benzaldehyde and formalin  solution are visible at
masses 107, 71, 33,  and 31.

      The chemical ionization mass  spectrum in 30 ppm of ethyl butyr-
aldehyde in methane reagent is  given in Figure 12.  The  principal ions
oroduced  from the ethyl butyraldehyde are  the P+l  ion at  mass 101  and
the fragment ion produced by  loss of water from the protonated parent
mass 83.    The ion at mass 59 may  be due  either to loss of C3H(,  from
the protonated parent or due to  a significant impurity of either propion-
aldehyde or acetone in the sample.   The other ions in the spectrum are
due principally to impurities residual from the previous samples.

      The chemical ionization mass  spectrum for 35 ppm  of iso-butyral-
dehyde in methane reagent is  given in Figure 13.    The principal ion
produced  from the iso-butyraldehyde is the P+l  at mass  73.  Fragment
ions due to loss of H2 and loss of water from  the protonated parent are
observable at masses 71 and 55 respectively.
                                  -32-

-------
     ,y
   I




 _ __!___
rP
           I

           _L
                 -t—
                          I
                 l-r
                                            -t-  o
                                               o
                                               M
                                               O
                                               «
                                               O
                                               oo
                                                       w
                                                   CQ
                                                       CO
                                                       CO
                                                       ri

                                                      s
                                               o
                                               rO
                                               O
                                               rj
                                                               V

                                                               >>

                                                               9)

                                                              2
                                                              "(3
                                                               c
                                                               ^  c
                                                              u  §
   u
   o

   °
 o _
 « "g
                                                               §15
                                                              rt
                                                                  v
                                                                  oo
                                                                  n)
 O

H! rt




 S jS
O
^H

CO




.^

h
          J-)
          •H
          CO
          d
          0)
          w
          -33-

-------
"Gi
Ci
              r
             7~

            ° \
            cc •
            o -(-
            CJ
            UJ
                    UJ
                    _l
                    
-------

^
V
^
-
<*
£
&

>-
fc
s
s




































\
c






































k
h
k
K
h>
1










































































































































"*
























































































































»


<5














i—i






















•^
hx



S)




































««



•




«••


=
—























O

Q
?

cc

r
CC
CD
Z

a:
a.
tr
Q -
O

LU
^
»


























«
D
Z
D
UJ
Z
CC

O
I
O
D
Z
_J
UJ
UJ
CJ '









1
;








































>*
_»
-I






















































— ^.
4-
u







































.


























o
(VJ



J
^,
^







•**>
^































)












o
«>•






--•"- "

































1






=d

























0
••
=3














































.1.—











































«^C£
•^



••














I
1
~T
•H*l








— •+

5
J -



j



I
j —


3
i
j

{
?
5 —

!

. I
1

:




o
Tj* 4)
- s
rt
o ^
-H ..,
"H
?
- i
^
o 2
-H n)
^< h
t^
-»j
!3
o CQ
2 >> *
?*» —LI
J3 ^
M ^
W a)
O ti f°
^ o a
«« a
'm' JH ^
0 H 'S ^ §
«n & tJrt
S 
^ W ^ w S
C/D O flj r«i
51 s-«
o^3 g 9
VO ^ -H .
^ ^> M
^q 0
• r-l -M
0 g ^
in °
I-H ^j
-H rt
-§a
§ §&
U rt
^J 4)
O tf
o •
(1)
o ?«
tvj M
»H

fa

 CO
 C
 4>
W

35-

-------
•V
_cL
 c^
  ...4 .

    v/i>
   -nh
:?V-
         "K
               -- -i  -t	
                   I   I

              L. |_..[_ _
              r- ,•

                           i_


                           j
                           4
                               	^-4_^-
                                 j  i J  i
                                 •4- -t—l	


                                      -4

                                               1—4-
                                                             o
                                                             ^
                                                             o
                                                             ro
                                                             o
                                                             i-H
                                                             O
                                                             o
                                                            §
                                                            s
                                                            o
                                                            in
                                                            o
                                                            •*
                                                                    CO

                                                                    CO
                                                                 W
                                                                          2£
                                                                         "7 £
                                                                          0 c
                                                                         .2 §
                                                                            u
                                                                      a,


                                                                      CO
                                                                      to
                                                                        «
                                                                      N  ®
                                                                     ._,  t>0

                                                                      C  ni
                                                                      O  a)
                                                                      B^
                                                                      (U -M
                                                                         0)



                                                                         tiO


                                                                        £
                         CO
                         C
                         0)

                        w
                         -36-

-------
      The chemical ionization mass spectrum for 270 ppm of propion-
aldehyde in methane reagent is given in Figure 14.   The  principal
ion produced from the propionaldehyde is the P+1 at mass 59.   A
small amount of the protonated dimer ion is observed at mass  117.

      The chemical ionization mass spectrum for 200 ppm of n-butyral-
dehyde in methane reagent is given in Figure 15.  The protonated par-
ent ion occurs at mass 73 and in this case the  fragment ion due to loss  of
water at mass 55  is larger than the protonated parent ion.  It  is un-
likely that the large peak at mass 59 i s due to fragments  of the benzalde-
hyde fragment ions but is probably either propionaldehyde residue left
from  the previous sample or due to a significant impurity of either
oropionaldehyde or acetone in the butyraldehyde sample.

      The chemical ionization mass spectrum for 60 ppm  of 2-pentanone
in methane reagent is  given in Figure 16.   The principal  ion produced
from  the 2-pentanone is the P+l at mass 87,   The other ions in the
spectrum are probably due to residual  impurities in the previous
samples.

      The chemical ionization mass spectrum for 40 ppm of iso-valeral-
dehyde in methane reagent is  given in Figure 17.   The  P+l  ion occurs
at mass  87 but the fragment ion due to  loss of water from the protonated
parent at mass 69 is larger in intensity.
                                   -37-

-------

u
^
•>.
•v.
u,
<:•
j
tr
--,
•^
o
5"
£
PN
vy






























—

--




—



































H
<

—

—





















































































































-

—



































*.
«,
•«•
u
C
a
V







































>^
•4
^
4
J
4
)
1
I
)









































-.-

r-





































-





































O
(VJ
X



-































































































-*
\


—



































>.













__,







































TT —







—

-

1



























^r3



	 i



—

—


U.'


































—

—


«f















,

•


•*iC
-^*^
—


—
^~
•"=

_
i


1


-

_


—


—


—














„_



^







O
^
^H

ro
i— i
S
tva ?
TJ
r£j ^0
0 ^ h
*-« 2  «1 S
<* N "
^< .2 bO
" — d OJ
O D
o - &
in ^
S «
•H C
art
X
^ S ^
43 0
U S

o •*
CO i-l
t>
1-,
S 5P

N


-38-

-------



I-U
^
T-
\

V
(^
o
cr(

T>






























<
t
(

f










._




J

:1


.._













S

•j
1










- -



.,


L-
-




































-
...


•


























—



__.


---


— —





:1
























-





---


























-5
/•>
>~s





,•
_






























K.





il


—

^,
























'!'
—



.__
— \

i 	




5M4



r~.
X




















~x



-


—


—

vsa
-



j
1




















*






__.

•


TS



^
























O
w -
->
Q
UMENTS
fe
2
BRUSH
^
g -
0
Q-
O -
O
i
I
t,
J
Q
C












1














i/:

z

_

u.


^
-1
>
-1
J
•4
f>
-«




















..









-- -





c
(V
X




















..









-





)



























-





























...




























t






















































"


























—



»«n


-_

































•TM^_


-




"*


-^














,-






^-_



^=;
J— »T-J


— i


^^-




-




















-
Us;
	

i..«
C,I2
** ^,

••••j
L-----,
Ks«*r,
>»i«-C
--
^-.


1 - •*


_


»~.


— i-
i

— -


—


—


	


-f

j


i


!



— —

I.
I
O
•"*•
I—I

o
CO
I— 1

o

'£
O <1) •
i— I T3 t~-
i-H ^s
41 £
c
0 ^|
-u
pq 0
£ -=• ° -
w co m r -v
•4J L/
rt c § 6
[T| ^3 J3 i^
§ 5 M S^
> W ^ Tl
(S ^ fi
11 rn *^
r~ w 'H " o
<; c J5 **
51 5 S -
<; c *>
o -— o "i
•*-* "tn
°5 §
.2 bO
„ C R)
° ° s
f— 1
nl D
CJ r^
° 's -^
u 2
o •
0)
Q 2
rj W>
•"-1

-39-

-------



^
X
^
R
?••
<
v^

c\>



































c
c
V





































.
•^
i-
J













































































































































































(







































^
X







































fs
c*





















1— 1
X



























































k=z^












"
—


=~
•— ^

~!


























_i







o

G
00
1-
UJ -
D
tr

§]
Sensitivity ^



































- <
:j
o

~ o_













































































— 7T




O































— .







-1






•\
O








































/
K


















-
J












































































?








































>^
**.















--a
_




































^=


-^ra








•











«




-=



-^
-


:
	 n—


"™f
«,









•>«_.








-4
T'

i
-4
-*-

i

|
!
• 11^
_j_
i







4-



i
i

1


O
i-H


t-H
•4-1
V
s ^
^ rt
tf
o v
-i d
r-H (Tj
J3
0 S
Of^
•-H _r|
-4-)
.— <
?

£ o
rt
'm d
o -"-" ^
00 0{ 'd ^
n >— I '
W P (vj
W w
^H ~ h
0^2 °
r^ P 4n .
'Z ^ d 00°
w .a g n
< S o -S
O ^ O 
-------

tf

£
s
^»
\JJ
"^
[*
0
>~
^


































^
r,
^
*-,








































_
b






.


















































































































































__





























































































































—







































































'












»•.

























































































































































\

































































































t~^










~
:-£
*









+
cv










f





































































'— ^.
c*








•1

































V.







o







































=

,
-








=





...


















*— «»



-s
—






-

I"L_


— "
-~


—




	
|
4.
J_


— 1-
i

— ~r


!








I
1
	


— .


^
i

T


r
	 [
!



§ «
•— i C
•4-*
o S
-H J3
• -4
£
s »
u
J3
1)
o 2
-H k
4>
^H
n)
^
r-t ™
— ' M
n «
o -2
«*H &
o d S
^ ^ § Z
«5 ^
0 M J5 0) OJ
m t— ' n
| « " o
§| £'o
Jz S rt in
o ^ ^ 5; _
w 'S « 'S
<; S o g
*g o -^ rt
o ^ rt »-(
vC ^ N 1-i
^ "2 o
C -!->
^ -
O — 1 4J
in ro ro
SS
(w

^ 0 P(J
•
r—

0
ro 0

M
o h




          o
          (M
 CO
 C
 D
-41-

-------
      The chemical ionization mass spectrum for 50 ppm n-valyralde-



hyde in methane reagent is given in Figure 18.   The protonated parent



ion is at mass  87 and again the fragment ion at mass 69 due to loss of



water from the protonated parent ion is more intense than  the P+1 ion.








      The chemical ionization mass spectrum for 55 ppm for 3-butanone



in methane  reagent is given in Figure 19.  The only significant ion pro-



duced from the 3-butanone is the P+1 ion at mass 73.








      The chemical ionization mass spectrum for 80 ppm of crontonaldehyde



in methane  reagent is given in Figure 20.  The only ion produced from



the crotonaldehyde is the P+1 ion at mass 71.   This measurement is a



repeat of the earlier measurement at higher concentrations of the croton-



aldehyde given in Figure  10 and represents the  mass spectrum obtained



under analytic conditions.








      The chemical ionization mass spectrum for 350 ppm  of acetaldehyde



in methane  is given in Figure 21.  The principal ion produced from the



acetaldehyde is the protonated parent at mass 45.    A small amount of



the protonated dimer ion  is observed at mass 89.








     The chemical ionization mass spectrum for 270 ppm  of acetone  in



methane reagent is given in Figure 22.   The protonated parent ion occurs



at mass  59 and no significant  fragment ions are observed.   A small



amount  of the protonated dimer ion is observed at mass 117 and a



fragment due to loss of water from the protonated dimer at mass 99.
                                  -42-

-------
 fci-
pr
-o -
IX
                r
i
                        o
                        rt
                        k
                            .-__!

                            i
                                  __t—j—i
                                     4
                                              i

                                             !
                                                0
                                                on
                                                o
                                                M
Q

n)
                                                          a

                                                          «> 2
                o
                o
                                                o  H
                                                   S m
                                                   P «!
                                                v£)
                                                O
                                                LO
                                                o
                                                   in o

                                                   to in
                                                   «> r-H
                                                   rt  >i
                                                   SI  O

                                                  '3  **
                                                   o  ^

                                                  -i  rt
                                                   rt  .j

                                                  .S  C
                                                     SOJ
                                                     GJD
                                                   I)  rt
                                                  .£  0)
                                                         oo
                                                          tuO
            X
                    01
                    d
                  -43-

-------

>
o
0
1
§























" "l
1
,--
—


















—
--




o
fL




















-





__.
—







—










__




-

--
—







—










_




.
-


-







—










—




—

—
—







—
ex










"



_..

—
..._


















— -
...

I



—
.__







>
•V














-

.._









—1









—


""•



—
....


















_..


-

ENTS DIVISION
5











—










*




i
!»










—










-















—








. —


-






-








—


or




•
-

























•U.U

j















—
•X



•"
-


ta^rr

n i rr

— .


i
1


-

— *•


—

j



|

1

I

i
_
1




o ^
* $
60
n)
o «
CU
0
O ^2
(M 4J
i—4 CU
O -P
i— 1 — '
CU
o
0 £j
o S
?
PQ
i
0 ^
0s L^
^ 0 .
CO **-< ^,
0(5 -3 S "
S w 5 9 M
00 CQ ^ K CU
*^ jj r^
^ w o-S
fi co a) H
t> rt p. 3
O x7 ^ rn *-r
fi ^ 2 w z
CO .-1 CO rj

-------
Vj,

>
-V-4
1
-Hi-
<
DC
o
0.
DC
O --
U
LU --
J_l
                                                        --4
                                                        J
                                                             O
                                                             ro
                                                             O
                                                             o
                                                             o
                                                             cr-
                                                                   to
                                                                   4-»
                                                                   • H
                                                                   a
                                                             oo
                                                                A  w
                                                                p  rt
                                                                w
                                                                w  •
                                                             o
                                                             vJD
                                                             O
                                                             in
                  o
                  ro
                  O
                  ro
                              0)
                              bo
                              n)
                              0)
                              0)
                             TJ
                              >N
                             XI
                              a)
                             T3
                             —1
                              rt
                              a
                              o
                             U
                              5  
-------

V
T>
lii

P

^
^
^














—
r






<
i
r-
1
0













—
—


^
'


--

-


1-

)
I
•)













-
—


/>
r
\







—
















—











—

















-•










—
















__











—




























_.














i

_
—










.-

















__




























—










-
























-


H
Z
BRUSH INSTRUME
2
I—
CC
0
Q_
CC
u
K
> J
LU
_J




















t/.
LJ
'S
CL
5
2
LJ
LJ
_J










































	










"
















—
—




























	







































_
-
r\
kj
















-:




























i










..
-
-











i.
















"










4-






•rt


i«-rn







_










r-
(>3



-
s
S




^_ •

'-"^



,_











-=




"*•

•

.
~a
~^Z3
	
t~--^

._
i
1


	
<
i
—
i
t

	 '







	
_







_


O
»— 1
O fl)
i— 1 rrt
4-J
s 1
™ *
(H
5
.-4
£
O
-H T3
JS
^5 'O
O "^ .
"• JS 00
o
S >H ^
op
"^ ^3
rt -3 g ^
„ w ^ g c
° m '-J ^3
co " -*-1 >
•^ to o K
i-^ »_H rn C5
° 2 S ^ o
W 0 ? 0
t/) ._< OJ "^
 O I^H T5
0 i c «
NO 
-------





10
-\
£


fo
^












r
0
,
r\


















v
i.
-
r
Ci














_
0





















"-,
-












































































































































































































































































































































































g

0
i- -
5
cc
2
I
C/5
CC
CO


cc
0
CL
CC
o

LU




























CO
z
1 o
UJ
z
CC

0
I
o
Q
z

_J























































































































































(
c







































s
^
^x


























"^

















c
-'








































i












































j-
f



















































r i





; 	 1

















»
















r — mf
££?
•


H
ti=
==

•v^
r

—













































Q
>— <

O
ro
-H

O
(VJ
—i

O
t-H
^H
o



o

CO
O M C
00 S ^
S w
S M
01-7. rt
r- ,2;
W u
S3 •?
o s§
<-J <^H ^>


                                                               V
                                                               (U
                                                               c
                                                               o
                                                               -4->
                                                               cu
                                                               o
                                                                   t-l
                                                                   CD
                                                               °o°
                                                              .-I  0


                                                               rt  in
                                                               n)
                                                               U
                                                              U  a


                                                                •
                                                              (M


                                                               cu



                                                               bo
                                                              •0-1

                                                              h
 CO
 C
 cu
in
            o
            ro
-47-

-------
      The chemical ionization mass spectrum for  140 ppm of acrolein



in methane reagent is given in Figure 23.   The principal ion in the



spectrum is the P+l at mass 57.   A very small amount of the dimer



ion at mass 113 is also observed.   No significant fragment  ions are



observed from the acrolein.







      The chemical ionization mass spectrum for a methane  blank  run



immediately following the preceding runs is given in Figure  24.   This



measurement was done virtually under the same conditions as the  pre-



ceding runs on the aldehydes.   The bottom trace on Figure  24 corres-



ponds to increasing the sensitivity by both increasing the electron  beam



current by a factor of six and increasing the recorder sensitivity by a



factor of ten.   This result shows the noise level of the instrument and



indicates that a few parts per billion of the aldehydes should be detect-



able at the highest sensitivity.







      Run No. 17 (Figure 25) was  carried out to assess the  sensiti/ity



of the chemical ionization method.   The source pressure was increased



for this run to 5 torr and the electron emission was  increased to 2 milli-



amps .   In this measurement a sample consisting  of 2 ppm of benzalde-



hyde  in methane was  used.   The signal for the M+l  ion (mass 107) was



600 millivolts per ppm of benzaldehyde.   The noise level was two  milli-



volts  rms.  This measurement indicates that a concentration of 10 parts



per billion of  benzaldehyde would give a signal-to-noise ratio greater



than 3.
                                   .48-

-------
Ci,
                                         rs
                                            rt

                                                                   o
                                                                   fM
                                                                   O
                                                                   o
                                                                   O
                                                                   o
                                                                   0
                                                                   oo
                                                                       CQ p


                                                                       5  »
                                                                           o


                                                                           o
                                                                   o
                                                                   in
                                                                   o
                                                                   m
                                                                   o
                                                                   M
 0)
 3
 ri
J3
4->
 4)

s
^3
 4->
• H
 ^

 c
«^H
 (1)
^H   •
 O tn
 ^H --H


^S


 si
 £  c
    3
 o rt
 D
 cx  -
w U
   o
                                                                                  CO
                                                                                    •O
 » uh
 

 g-"
M ^
-i  rt

 II **
•H  «
 d  <"
 S  bo
 cu  nJ

U (U
                                       o
                                       CM
                                       X
                            CO
                            C
                           -49-

-------
f \~i-i
   -_l_|__|_4i-
                       t>
                             .4_


                                      i

                                      i
                                                                         TJ
                                                                         at



                                                                         n)

                                                                         ffl

                                                                         0)
                                                                         0)

                                                                        w
II
si
N  p


I  §

u   .


l-o
•£ in


(M
0)

bo
• -H
PR
              w
              C
              0)
                         o
                         04
                                            o
                                            o
                                            X
                                  -50-

-------
                                                            to  c
                                                            <  o
                                                                     14

                                                                     U

                                                                     O
                                                                     O
                                                                     •o
                                                                     X

                                                                     v

                                                                     2
                                                                     13
                                                                     M

                                                                     «
                                                                     w

                                                                     I*
                                                                     O
                                                                      y
                                                                      v
                                                                     w
                                                                      «»  ^)


                                                                     2 1

                                                                     II
                                                                     •- <;

                                                                      1  c
                                                                     .5  3
                                                                      c «
                                                                      o
                                                                     »— (  »
                                                                     _ u
                                                                      flf •
                                                                      u o
                                                                     .- m

                                                                      E -
                                                                     U  rt
                                                                      U
                                                                      IH


                                                                      60
 B
 V
tn
                               o
                               o
                               rj
                               X
             -51-

-------
      In the absence of interference and fragmentation it is expected



that the sensitivity for the other aldehydes and ketones should be



comparable.   The larger saturated aldehydes; butyraldehyde,  valer-



aldehyde, and ethyl butyraldehyde,  all show substantial fragmentation



of the M-f-1 ion  produced by proton  transfer from methane reagent ions.



These fragmentations, which involve loss of hydrogen molecule and



loss of water molecule as the principal reactions, do not occur to any




appreciable extent for the isomeric ketones studied.   The fragment



ions corresponding to loss of H2O and H2 may be used to distinguish



between the aldehydes and ketones.  These fragmentations cause the



sensitivity, as  determined from the M+l ion intensity, to be  somewhat



lower for the larger saturated aldehydes than for the other compounds



studied.








      The saturated aldehydes and  ketones do not  suffer from inter-



ference with the hydrocarbons present.  The oxygenated compounds



give ions  orincipally at the parent  mass plus  1, while the hydrocarbons



of the same molecular weight yield principal  ions at the  parent mass



minus 1.   The protonated alkanes  larger than ethane are apparently



not stable and the  protonated species have not been observed in the



mass spectrometer.   On the  other hand, the unsaturated aldehydes,



e.g. acrolein and  crotonaldehyde.are  interfered with by the C4  and C5



hydrocarbons respectively.   Butyl ions at mass 57 may be produced



by proton transfer to the butenes and by hydride ion transfer from the



butanes.     These reactions produce ions of the same nominal  mass



as the protonated acrolein.
                                   -52-

-------
      It appears likely that a different reagent will be required for



the measurement of the unsaturated aldehydes in engine exhaust.



Additional experimental data  is needed to choose an appropriate re-



agent.  One possibility is ammonia which will produce NH4  as the



principal  reagent ion.   This  ion may transfer a proton to acrolein



and not to the butenes.   However, NH4  will certainly not react with



formaldehyde.   Thus it may  be necessary to do the analysis in two



steps; using methane to measure the saturated aldehydes and using



a different reagent to measure the unsaturated aldehydes.  Further



experimental  work will be undertaken to establish the reagent and



operating conditions with which acrolein can be analyzed in engine



exhaust by using the chemical ionization technique.
      The results of the present work indicate that the chemical ioniza-



tion direct reading mass spectrometer can be used to monitor individual



saturated aldehydes and ketones  in engine exhaust in the concentration



range from 10 parts per billion to 300 parts per million.   Further work



is  required to fully establish the sensitivity,  accuracy, and specificity



of  this technique in the application to  routine exhaust measurements.



Additional work is also required to extend the technique to unsaturated



aldehydes and to other  components of interest.
                                -53-

-------
                    A PPENDIX






SELECTED CHEMICAL IONIZATION BIBLIOGRAPHY
                       -54-

-------
                       SELECTED CHEMICAL IONIZATION BIBLIOGRAPHY
 1.  F. II.  Field,  "Chemical loni/ation Mass Spcc-
    trornelry, " Accounts of Chemical Research 1,
    42(1968).

 •2.  M. S. 15. Munson  and  F. II.  Field, "Chemical
    loni/ulion  Mass Spectromelrv.  I. General In-
    troduction,"]. Am. Chem. Soc. 88, 1621 (1966).

 3.  M. S.  B. Munson and  F. H.  Field, "Chemical
    lonization  Mass Spectromelry.   Jl.   Esters,"
   ,1. Am. Chem. Soc. 88, 4337 (1966).

 4.  F. II. Field, M. S. B. Munson  and D. A. Becker,
    "Chemical  lonization  Mass Spectrometry, Pa-
    raffin  Hydrocarhons,"  Advances in Chemistry
    Series, No. 58, American  Chemical Society,
    Washington, D. C. 1966, p. 167.

 5. M. S. B. Munson and F. II. Field, "Chemical
   lonizalion  Mass Spectrometry.  IV.  Aromatic
   Hydrocarhons," J.  Am.  Chem. Soc. 89, 1047
   (1967).

 6. F. H. Field and M. S. B. Munson, "Chemical
   lonization Mass Speclrometry.  V.Cycloparaffins,"
   J. Am. Chem.  Soc. 89, 4272 (1967).

 7.  F. H. Field.  "Chemical  lonization Mass Spec-
    trometry. VI.  CY HJJ Isomers.  Toluene, Cyclo-
    heptalriene, and Norbornadiene," J. Am. Chem.
    Soc.  89, 5328(1967).

 8. F. H.  Field,  Peter Hamlet and W.  F. Libhy,
   "Chemical  lonization Mass Spectromelry, VII.
    Reactions  of  Benzene  Ions  with Benzene,"
   J. Am. Chem. Soc. 89, 6035 (1967).

 9. F. H. Field,  "Chemical lonization Mass Spec-
   trometry. VIII. -\lkencs and Alkynes," J. Am.
   Chem. Soc. 90, 5649 (1968).

 10. F. II. Field,  "Chemical  lonizalion Muss Spec-
    Irometry. IV Temperature and Pressure Studies
    with  Benzyl  Acetate and  l-Amyl  Acetate,"
    J. Am. Chem. Soc. 91., 2827 (1969).

II.F. 11. Field. "Chemical lonizalion Ma.-s Spec-
   lromelr\.  \.  Temperature Simile* with Sub-
   stituted  Ben/\l Acetates," J.  Am. Chem. Soc.
   9J, 6331(1969).

 12. I). P. Weeks and F. II. Field, "Chemical loni-
    y.alion  Mass  Speclronielr). \l. Reaction*  of
    Melho\\ mclli\ I  Formate and Melho\\ methjl
    Acetate \\ilh  Methane and Isoliutane," J. Am.
   Cliem.Soe.92. 1600(1970).

 13. F. II.  Field,  "Chemical loni/ation Mass Spec-
    tromelry. \ll. Mcohols," J.  Am. Chem. Soc.
    92.2672(1970).
 14. F. II. Field and M. S. B. Munson, "Reactions of
    Gaseous Ions. XIV. Mass Speetromelric Studies
    of  Methane  at Pressures to 2  Torr," J. Am.
    Chem. Soc. 87, 3289 (1965).

 15. M.  S. B. Munson and F. II. Field, "Reactions of
    Gaseous Ions,  XV. Methane  +1% Fthane and
    Methane + 1%  Propane," J. Am. Chem. Soc.
    87,3294(1965).

 16. M.  S. B. Munson and F. H. Field, "Reactions of
    Gaseous Ions, XVI. Effecls of Additives on Ionic
    Reactions in  Melhane," J. Am. Chem. Soc. 87,
    4242 (1965).

 17. M.  S. B. Munson and F. H. Field, "Reactions of
    Gaseous Ions. XVII. Melhane  + Unsaturaled Hy-
    drocarhons,"  J.  Am.  Chem.  Soc.  9]., 3413
    (1969).

 18. J. H.  Fulrell and T. 0. Tiernan, "Ionic  Re-
    actions of  Unsaturaled Compounds.  I. Poly-
    merizalion  of Acetylene," J. Phys. Chem. 72,
    158(1968).                             ~~

 19. T.  0. Tiernan and J.  H.  Futrell, "Ionic  Re-
    actions in  Unsaturaled Compounds.  II.  Eth-
    ylene,"J. Phys. Chem. 72, 3080 (1968).

20. F. P. Abramson and J. H.  Futrell, "Ionic Re-
    actions in  Unsaturated Compounds.  III.  Pro-
    pylene  and  ihe Isomeric  Butenes." J.  Phys.
    Chem. 72, 1994(1968).

21. B. M. Hughes, T. 0. Tiernan and J.  F. Fulrell,
    "Ionic Reaclions  in Unsaluraled Compounds.
    IV. Vinyl Chloride," J. Phy. Chem.  73, 829
    (1969).

22. J. H. Fulrell, T. 0. Tiernan, F. P. Abramson,
    C. D. Miller,  "Modification  of a Time-of-Flight
    Mass  Spectrometer for Investigation  of Ion-
    Molecule Reactions at Elevated Pressures," Rev.
    Sci.Instr. 39, 340(1968).

 23. E. Gelpi and J. Oro, "Chemical lonizalion Mass
    Speclromelry of Pristane, " Anal. Chem. 39,388
    (1967).

24. II. M. Fales, G. W. A. Milne and M.  L. Vestal,
    "Chemical  lonizalion  Mass Speclromelry  of
    Complex Molccuh-s," J. Am. Chem. Soc. 91,
    3682 (1969).

 25. II.  M. Fales,  II.  Lloyd and  G. W. A. Milne
    "Chemical  lonizalion  Mass Speclromelry  of
    Complex  Molecules.  II. Alkaloids,"  J.  Am.
    Chem. Soc. 9>2, 1590(1970).

                                        (conl.)

-------
26. H.  Ziffer, H. M. Fales, G. W. A. Milne and F.
    II.  Field, "Chemical lonizalion  Mass Speclro-
    metry of Complex Molecules. 111. Structure of
    the Photodimers of Cyclic a—/3 Unsaluraled
    Ketones," J. Am. Chem. Soc. 92, 1597 (1970).

27. J. H. Futrell, F. P. Abramson,  A. J. Bhatta-
    charya and T.  0. Tiernan, "Proton  Transfer
    Reactions in  the Simple Alkanes:   Methane,
    Ethane, Propane, Butane, Pentane, and Hex-
    ane," J. Chem. Phys. 52, 3655 (1970).

28. H.  M.  Fales,  G. W. A. Milne and T. Axenrod,
    "Identification  of  Barbituates  by  Chemical
    lonization Mass Spectrometry," Anal. Chem.
    42, 1432 (1970).

29. G.  W.  A. Milne, T. Axenrod and H. M. Fales,
    "Chemical  lonization Mass Spectrometry of
    Complex Molecules. IV. Ainino Acids," J. Am.
    Chem. Soc. 92, 5170(1970).

30. T.  0. Tiernan  and J. H. Futrell, "Studies in
    Chemical lonizalion Mass Spectrometry. I. Re-
    actions of Selected Proton and Hydride Trans-
    fer  Reagents  with Hexane Isomers," Rev. Sci.
    Instrum. (in press).

31. L. Wojcik and J. H. Futrell, "Studies in Chem-
    ical lonization Mass Spectrometry. II. Aliphatic
    Alcohols," Biochem. Biophys. Res. Commun.
    (in  press).
32. D. M. Schoengold and M. S. B. Munson, "A
    Combination of  Gas  Cbromalography  and
    Chemical   lonizalion  Mass  Spectromrtry,"
    Anal. Chem. J_2, Dec. 1970 (in press).

33. D. Beggs,  H. M. Fales, G.  W.  A. Milne and
    M. L. Vestal,  "A Chemical lonization Mass
    Spectrometer  Source," Rev.  Sci.   lustrum.
    (in press).

34. A. A. KiryushUn, H. M. Fales, T. Axenrod, E.
    J. Gilbert and G.  W. A. Milne, "Chemical loni-
    zation Mass Spectrometry of Complex Mole-
    cules.  VI. Peptides,"  J.  Am.  Chem. Soc.
    (in press).

35. D. P. Beggs and F. H. Field, "Reversible Re-
    actions of Gaseous Ions. I. The Methane-Water
    System," J. Am. Chem. Soc., to be published.

36. D. P. Beggs and F. H. Field, "Reversible Re-
    actions of Gaseous Ions. II. The Propane-Water
    System," J. Am. Chem. Soc., to be published.

37. F. H. Field and D. P.  Beggs, "Reversible Re-
    actions  of Gaseous  Ions. III. Studies with
    Methane  At P = 0. 1 - 1.0 Torr  and T = 77-
    300° K," J. Am. Chem. Soc., to be published.

-------
    THE  FOLLOWING  PAGES  ARE  DUPLICATES  OF
ILLUSTRATIONS  APPEARING   ELSEWHERE  IN  THIS
   REPORT.    THEY  HAVE  BEEN  REPRODUCED  BY
  A  DIFFERENT  METHOD  SO  AS  TO  FURNISH  THE
           BEST  POSSIBLE  DETAIL  TO  THE  USER.
:.;:'%&J. ;« .•-*:•,.'.-,*:, -
                                     ..-^V-v '-;^^^'^^2i^-^3*^-;1:^>T,,v,:.2^

-------
 ti  CO
 ni  nJ

 ^s
 CD
3
 ni
 4)  0
^  O
w •-


 S  S


   '
   U
0) i.

XI W


"" ^
m CO
 to  O
-u -*->

 c  o
 
-------
                     .u
                 .u.u
                 •u.u
Figure 4.  A Photograph of the Completed Prototype System.




The electronics cabinet on the left contains all of the readout and

control systems and the cart on the right contains the direct reading

mass spectrometer with chemical ionization source and associate

vacuum systems and sample inlet systems.
                            -18-

-------