CHIEMICAL

RESERRCDI.
DEVELOPMENT ri
ENGINEERINGI
CENTER

ELECTRICAL PROPERTIES OF ATMOSPHERIC MOIST AIR:

A SYSTEMATIC, EXPERIMENTAL STUDY

Hugh R. Carlon
LI) U.S. Army Fellow
RESEARCH DIRECTORATE

September 1988

DTIC

U.S. ARMYT

MUNITIONSV
CHEMICAL COMMAND

Aberdeen Proving Ground, Mariand 21010-5423

'20T
I w

Disclaimer

The findings in this report are not to be construed as an

official Department of the Army position unless so designated
by other authorizing documents.

Distribution Statemient

Approved for public release; distribution is unlimited.

 UNCLASSIFIED
SECURITY CLASSIFICATION OF THIS PAGE

REPORT DOCUMENTATION PAGE

la REPORT SECURITY CLASSIFICATION lb RESTRICTIVE MARKINGS
UNCLASSIFIED
2a SECURITY CLASSIFICATION AUTHORITY 3 DISTRIBUTION/ AVAILABIUITY OF REPORT

2b DECLASSIFICATION/DOWNGRADING SCHEDULE Approved for public release;

distribution is unlimited.
4 PERFORMING ORGANIZATION REPORT NUMBER(S) S. MONITORING ORGANIZATION REPORT NUMBER(S)

CRDEC-TR-88059
6a NAME OF PERFORMING ORGAN'ZATION 6b OFFICE SYMBOL 7a. NAME OF MONITORING ORGANIZATION
(If applicable)

CRDEC SMCCR-RSP-P
6c ADDRESS (City, State, and ZIPCode) 7b ADDRESS (City, State. &ad
ZP Code)

Aberdeen Proving Ground, MD 21010-5423

Ba. NAM.E OF FUNDING/SPONSORING 8b. OFFICE SYMBOL 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER
ORGANIZATION (If applicable) U.S. Army Science and Engineering
CRDEC SMCCR-RSP-P Fellowship
Sc ADCESS (Cty, Ct, . ::' Code) 10 SOURCE OF FUNDING NUMBERS
PROGRAM PROJECT TASK WORK UNIT
ELEMENT NO. NO. NO. ACCESSION NO
Aberdeen Proving Ground, MD 21010-5423
11 TITLE (Include Security Classfication)
Electrical Properties of Atmospheric Moist Air: A Systematic,
Experimental Study
12 PERSONAL AUTHOR(S)
Carlon, Hugh R., U.S. Army Fellow
13a TYPE OF REPORT 13b TIME COVERED 14. DATE OF REPORT (Year, Montt, Day) S. PAGE COUNT
Technical FROM 86 MayTO87 May 1988 September 31
16 SUPPLEMENTARY NOTATION

17 COSAT CODES 18 SUBJ;Z" TERMS (Continue on reverse if necessary a iDentify by block number)
FIELD GROUP SUB-GROUP Atmospheric electricity, Field strengths3
04 01 Electrical conductivity, Surface films. V,,.
Static electricity, (Continued on revers~ k
19 ABSTRACT (Continue on reverse if necessary and identify by block number)

Systematic new measurements of the electrical properties of atmospheric moist

air in a closed container near sea level are presented. Data were taken
using a vapor electrical conductivity cell of new design. The cell consisted
of 40 parallel square plates, each 26.5 cm 2 -, and separated by 0.66 cm. The
cell insulators were isolated from the cell plates by use of a technique that
completely eliminated insulator leakage as a source of experimental error.
Thus, very sensitive measurements could be made in still or moving air even
at near-saturation humidity. Measurements were made over a wide rang- of
humidities and electric field strengths ranging from near zero to 3 kV/cm. '

(Continued on reverse)

20 DISTRIBUTION / AVAILABILITY OF ABSTRACT 21. ABSTRACT SECURITY CLASSIFICATION

OUNCLASSIFIED.,JNLIMITED 0
SAME AS RPT [-' OTIC USERS UNCLASSIFIED
22a NAME OF RESPONSIBLE INDIVIDUAL 22b TELEPHONE (include Area Code) 22c. OFFICE SYMBOL
SANDRA J. JOHNSON (301) 671-2914 SMCCR-SPS-T
DD FORM 1473,84 MAR 83 APR edition may be used until exhausted SECURITY CLASSIFICATION OF THIS PAGE
All other editions are obsolete UNCLASSIFIED

UNCLASIFIE
 UNCLASSIFIED
SECURITY CLASSIFICATION OF THIS PAGE

18. SUBJECT TERMS (Continued)

Water vapor
Humidity
Atmosphere
Ions

19. ABSTRACT (Continued)

Many new insights into the behavior of moist air in electric fields were
gained, including: (1) a very steep equilibrium dependence of conduc-
tivity (ion or charge carrier content) upon humidity; (2) "hysteresis"
effects; (3) increasing complexity of behavior at field strengths
greater than a few hundred volts per centimeter; and (4) regions where
the conductivity of near-saturated air was constant over a range of
field strengths. All results were consistent with the view that ions
are produced in moist air from thin water films on conductive surfaces
S.. between which an electric field exists. This effect could contribute
'Si significantly to the earth's global current.

.%
I_r CO-'Y
INS,: rro
4

ul.J
; ' ,. -. : :) . .. .

% %. %

%u %

:.'. UNCLASSI FI ED

. ..
.. .. .. ...
.. . .. :,*, ,-*- --.
,, ,,," . . . . . ". . . .. ""S
."."." - - ..-.
" .
-'- . ,* '...-
'. ^.
 PREFACE

The work described in this report was authorized under a

U.S. Army Science and Engineering Fellowship. This work was
started in May 1986 and completed in May 1987.

The use of trade names or manufacturers' names in this

report does not constitute an official endorsement of any
commercial products. This report may not be cited for purposes
of advertisement.

Reproduction of this document in whole or in part is

prohibited except with permission of the Commander, U.S. Army
Chemical Research, Development and Engineering Center,
ATTN: SMCCR-SPS-T, Aberdeen Proving Ground, MD 21010-5423.
However, the Defense Technical Information Center and the
4q National Tpchnical Information Service are authorized to
reproduce the document for U.S. Government purposes.
This document has been approved for release to the

public.

U* -
 Blank

44

J% 4
 CONTENTS

Page

1. INTRODUCTION ...................................... 7

2. APPARATUS AND METHOD .............................. 10

3. RESULTS AND DISCUSSION ............................ 12

4. CONCLUSIONS ....................................... 29

LITERATURE CITED .................................. 31

LIST OF FIGURES

1. Diagram Showing How Cell Insulators Are Kept

Warm and Perfectly Dry by Hair Dryer in Duct
or "Saddle" Outside Box ........................... 9
2. Schematic Diagram of Electrical Measurement
Circuit for Cell 9........

3. Ion or Charge Carrier Population per Cubic

Centimeter of Vapor, Icc, vs. Saturation
Ratio (s = %RH/100) . ............................... 15
4. Data Plots of Voltmeter Reading, Ev (Figure 2),

and Cell Current from Which Average Moist Air

Ion or Charge Carrier Populations per Cubic
Centimeter between Cell Plates Can Be Calculated
(Eb = 0.4 kV) ...................................... 16

5. Data Plots of Voltmeter Reading, E v (Figure 2),

and Cell Current from Which Average Moist Air
Ion or Charge Carrier Populations per Cubic
Centimeter between Cell Plates Can Be Calculated
(Eb = 1.0 kV) ..................................... 17
6. Data Plots of Voltmeter Reading, E v (Figure 2),

and Cell Current from Which Average Moist Air

Ion or Charge Carrier Populations per Cubic
Centimeter between Cell Plates Can Be Calculated
* (Eb = 1.5 kV) ..................................... 18

7. Using Data Like Those in Figures 4-6, These Curves

Were Constructed for Near-Equilibrium Conditions
K 'with Bias Voltage Applied Continuously to the
Large Cell ......................................... 20

P
 8. From Figure 7, These Curves Were Constructed
for Near-Equilibrium Conditions with Bias
Voltage Applied Continuously to the Cell ............. 22

9. Data Taken for Moist Air at 24 OC, Fan On,

Continuous Bias Voltage, Showing Departure
from Linear (Ohm's law) Behavior at Higher
Bias Voltages ...................................... 23

10. Data Taken for Air at 23 *C Being Humidified

Using Ultrasonic Nebulizer, Fan Off, Continuous
Bias Voltage Decreasing for Readings ................. 24

11. Continuation of Data in Figure 10 for Air

Near Saturation, "Drying Down" Without Heating,
Fan Off, Continuous Bias Voltage Decreasing
for Readings ....................................... 25

12. Data Taken for Air at 27 *C Being Humidified to

"Hard" Saturation by Steam from Boiling Water,
Fan Off, Continuous Bias Voltage ...................... 26

0 m" %

* :s*".

6
 ELECTRICAL PROPERTIES OF ATMOSPHERIC MOIST AIR:
A SYSTEMATIC, EXPERIMENTAL STUDY

1- INTRODUCTION

This report presents rec-lts of a detailed, systematic

study of the ion content and electrical properties of moist air
at atmospheric pressure. Electrification of moist air by evapo-
rating liquid water, particularly if the evaporation is energetic
so that a finely divided fresh surface is produced, has attracted
the interest of theoreticians and experimentalists for over two
centuries. In 1782, Volta may have been the first to suggest
~1
that electrical charges are liberated by evaporating water, and
Faraday wrote in 1844 that jets of escaping wet steam can produce

strong electrification. 2 Lenard (1892) found that the splashing

of water produces charging, 3 while Simpson (1909) 4
and Nolan and
5
Enright (1922) observed charging from the violent breaking of
droplets, which depended to some extent upon dissolved sub-
stances, but they could not measure electrification due to the
gradual production of a new water surface. Blanchard (1961)6
showed that ions can be generated by bursting water bubbles.

Carlon (1980) first reported the very large ion or

charge carrier populations that can be generated in closed con-

tainers by boiling water. 7 Data were taken using large electrical

conductivity cells of a new kind, which used compensation for
insulator leakage currents. The basic design of these cells
is being used in the present work but with further, important
refinements. Measurements were extended to conditions of simple
8
evaporation of stagnant water by Carlon, and it was shown that
boiling water is not necessary to produce very large ion popula-
tions in moist air, although boiling did produce larger popula-
tions than did simple evaporation. This work showed that the
ions as measured between the cell plates were present in equilib-
rium populations that depended strongly upon relative humidity
(RH) and, to some extent, upon the way in which evaporation
(humidification) was carried out. For conductivity cells smaller
than those described below and elsewhere by Carlon, the conduc-
tivity of the water ions in moist air between the cell plates and
that due to leakage through thin films of water that condense on
7-9
insulator surfaces can become comparable. Both instances of
conductivity show a very steep dependence upon humidity that is
nearly identical for the insulators and for the moist air itself.

Moore discussed the insulator effects, 1 0 but the conduc-

tivity of moist air was not understood at that time. Investiga- 9
tions with better insulators and improved mechanical cell designs
confirmed that a new design that effectively eliminated the
insulators as a source of experimental error was required if

7 7

-- ~~~~~~~~~~~~~~~! 'r'. W---......................~- -

 ' !

unequivocal data were to be obtained, giving true ion populations

in the vapor, particularly in conditions approaching saturation
at high temperatures. This design, which has been proven in the
present work. will be discussed in detail in this report.

Measurements were made at electric field strengths

E(V/cm) ranging from near-zero to 3 kV/cm, thus including regions
where the electrical conductivity of moist air is linear with E
(i.e., where Ohm's law is obeyed), and those where the conduc-
tivity is decidedly nonlinear. Measurements were carried out in
* still and moving air over a wide range of RH and a limited, near-
ambient range of temperatures. The new cell is shown schemati-
cally in Figure 1, and the series electrical measurement circuit
is shown schematically in Figure 2. The supply or "bias" volt-
age, Eb, provided by one of two regulated power supplies, was
thus divided between the cell voltage Ec.
C and the voltage E read
by a Schlumberger Model SM-5228 vacuum-tube voltmeter (VTVM) that
'4 -,

had a constant electrical resistance of R = 11 M (1.1 x 10' 4)

regardless of the voltage range selected. Thus:
S
E = + E (la)
b V c
* or
E =E - E . (lb)
c b v
Since the electric current i in a series direct current
(dc) circuit is everywhere
amperes can always be given the
by: same, the cell current ic in

7 -8
ic = E v/R
v = Ev /1.1 x 10 vv = 9.1 x 10 E . (2)

Equation (2) can be used to calculate the cell current,

but in this report the proportional voltage Ev will always be
used in the equations and figures because:
a. It is mucti more convenient to express E as read
v
directly from the VTVM than to convert it to cell current [see
Equation (2)].

b.
The ratio Ev/Eb describes some very interesting
properties of moist air, as will be shown.

Because both cell current i and cell voltage E were measured,

c c
the electrical power dissipated per cubic centimeter of vapor in
the cell could be calculated accurately for given operating con-
ditions. The vapor electrical conductivity cell described in the
next section of this paper was operated at E b = 0 to 2 kV, thus
producing electric field strengths of E Eb/0. 6 6 = 0 to 3 kV/cm.

< <
6,
 01

END BLOWER

Figure 1. Diagram Showing How Cell Insulators Are Kept

Warm and Perfectly Dry by Hair Dryer in Duct
or "Saddle" Outside Box

11 MEGOHMS

II E

SBA SOURCE~

Figure 2. Schematic Diagram of Electrical

Measurement Circuit for Cell

9
 2. APPARATUS AND METHOD

A Lucite box 105 cm long by 35 cm wide by 46 cm high

contained the following apparatus:
a. A large vapor conductivity test cell or capacitor
(Figure 1)

b. Probes to measure dry bulb temperature and dew

point

c. A 10-cm-diameter muffin fan that provided air

circulation at about 1 m/s when desired

d. A 600-cc Pyrex beaker containing a 500-W immersion

heater and three plastic tubes that could be used to deliver cool
4 fog into the box from an external ultrasonic nebulizer, liquid
water into the beaker for boiling, or liquid water onto the floor
of the box for evaporation, respectively.

A Yellow Springs Instrument Co. (YSI) Model 91 HF-calibrated pre-

cision electronic dew point hygrometer was connected to the
probes and provided a direct meter readout of dew point and air
temperature in the box. It gave precision readings with the fan
on or off, after stabilizing to one or the other condition for
3-5 min. The muffin fan was a Dayton Model 100. The external
ultrasonic nebulizer, a medical instrument, was a Bennett Model
US-l, which provided energy control settings on a scale from
0 to 10 to determine the fog output rate. Lucite was chosen for
the apparatus wall material because it had been used for many
.'" 7-9
years in earlier work with no detectable difficulties, and it
allowed visual examination of water vapor condensation on the
walls that were often wet during experiments except beneath the
heated saddle as discussed below. A 10-cm hole in the end of the
box opposite the cell allowed access for tubes and wires. The
hole was sealed during experiments that were carried out at
atmospheric pressure. The box itself was sealed with rubber
gaskets and wing nuts on threaded studs around two access panels
that were used to remove or service the large cell or to clean
*the box.

The cell consisted of 40 square aluminum plates, each

26.5 cm square with an average separation of 0.66 cm between the
plates. Alternate plates were electrically connected to pairs of
threaded stainless steel rods that carried the weight of the
plates and to which leads were connected for conductivity read-
ings. The steel rods passed through small (1.3-cm diameter)
cleararnce holes in the walls of the box to the insulators outside
the box. The insulators were vertical strips of polytetrafluoro-
ethylene (PTFE) 1.2 cm thick by 2.5 cm wide by 39 cm long,
machined to sharp points at their bottom ends (Figure 1). The
points rested on PTFE shims that were used to center the steel
rods in the clearance holes. There were two sets of 20 plates

10
 each, carried on their own steel rods and insulators, that were
intertwined but otherwise not in direct p,,ysical contact. Thus,
the two sets of plates could be adjusted horizontally by moving
the insulators and vertically by using the shims to avoid contact
between the rods and clearance holes and between the two sets of
nested plates.

The insulators were enclosed in a removable duct or

"saddle" (shown by cross-hatching in Figure 1) that was used to
control their environment and, thus, to keep them dry during
experiments. This was the novel feature of this apparatus design
*'-" 7-9
compared to earlier designs, and it yielded unequivocal
experimental results. Since the same water surface films
condense on the cell plates and on unprotected insulators, any
insulator leakage could confuse experimental results because it
has the same functional dependency on temperature and water vapor
partial pressure as the electrical conductivity of the moist air
between the cell plates. By eliminating the insulator effects
altogether, as in the present work, the true vapor conductivity
can be measured.

The saddle containing the insulators was 32 cm wide and

5 cm thick along the side walls of the box, where the insulators
were housed and the steel rod clearance holes were located. At
its top, the saddle 1had a plenum with a height of 10 cm above the
top of the box, and a hole was provided at its top center to
mount a two-heat ('00/1200 W) domestic hair dryer. The 600-W dryer
setting proved adequate for all experiments. In the plenum were
a Cole-Parmer Instrument Co. (CP) Model 3310-40 certified temper-
ature and RH indicator and a precision CP Model 3310-20 RH
indicator. In operation, typical conditions monitored in the
plenum with the dryer operating at 600 W were 60 0 C and <15% RH
(the lower limit of the RH indicator). The dryer produced a
slight overpressure in the saddle, thus preventing the escape of
moist air from the rod clearance holes, further insuring that
moist air in the box during experiments did not reach the saddle-
housed insulators to confuse the vapor conductivity data. The
bottom of the saddle on both sides of the box was left open,
allowing hot air from the dryer to flow down over tlie insulators,
across the insulator points and the PTFE shims, and back into the
% atmosphere. An electronic thermometer (Electromedics Incorpo-
rated, Model M-99) monitored the air temperature at the insulator
points and shims. Readings of 50 0 C were typical here. In most
experiments, the standard practice was to maintain a differential
of at least 6 0 C between the temperature at the insulator points
and shims in the saddle and the dew point in the box although
differences of 20 C or more were common. But even if the saddle
dry air and box dew point temperatures were permitted to approach
one another, insulator leakage currents were not measurable
because of the extremely conservative design of the apparatus.
This could be demonstrated during the experiments, since the
current through the cell was not changed by starting or stopping
the hair dryer.

11

0As.
 K:-
W.
- _-11 - 1 T- W.W;WT. .-- . . . . ..

The water ion or charge carrier populations in moist air

between charged plates (electrodes) are very strongly dependent on
RH, as will be shown in the discussion of experimental results.
The ne..,= design provided the experimental conditions needed to
study the electrical properties of moist air with minimal error.
The cell was used over a wide range of humidities and bias
voltages d-wn to zero. These were conditions yielding relatively
low cell current densities, thus requiring excellent insulators.

The cell was wired in series (Figure 2) with either of

two dc-regulated power supplies [0-400 V dc Shlumberger
Model SP-2719 or 0-2000 V dc (2 kV) Brandenburg Model 475R] and
the VTVM. Under most experimental conditions, results revealed
' that when the supply or bias voltage was kept at 400 V dc or
less, plots of data for E versus Eb were perfectly linear, i.e.,
Ohm's law was obeyed. Thus, many humidification experiments were
conducted using the 400-V dc power supply, which gave linear
behavior for singly charged water ions except under extremes of
humidity, as will be discussed below. A variety of experiments
could be performed with a large cLll and its apparatus; humidifi-
cation could be carried out by using the ultrasonic nebulizer, by
boiling, or by evaporation of standing water in the box. Drying
* could be effected either by heating (i.e., by continuing to run
the dryer after humidification had ceased) or simply by allowing
drying to occur over many hours under ambient conditions. The
voltage Eb could be kept constant or varied over the cange of
0-2000 V dc (0-2 kV), producing in the latter case very interest-
ing data plots of E v versus E b* Measurements could be made in
either still or moving air. Operation of the fan produced
negligible ion populations in the moist air.

3. RESULTS AND DISCUSSION

The current density in A/cm 2 (j) for water ions in moist

air between the plates of a cell similar to those used in these
experiments is given by

j=I cc eNuE (3)

where
n 3
I the number of ions per cm of vapor at any
cc instant

e the value of the electronic charge (1.6 x

-19 C)

N the number of charges per ion

u the ion mobility

E the electric field strength (V/cm)

12
 6

Except where noted here, N is taken as unity and u as 1 cm 2 /V s,

which are typical parameters for singly charged small cluster
ions of perhaps 10-12 cr more water molecules such as are almost
certainly being measured in this work. For these measurements
of electrical conductivity in moist air the dc resistance or
reciprocal conductance of the cell in ohms, R , is
related to Icc by the equation

cc (4)
c

where L is the plate spacing in centimeters and A is the total

plate area of the cells in square centimeters. The large cell
consisted of 40 square metal plates, each 26.5 cm square,
separated by an average spacing of 0.66 cm. Thus, A = (26.5) x
2
*., 39 = 27,388 cm 2
For the series circuit in Figure 2,

R = -b1 (5)

where the constant resistance of the VTVM was R v = 11 MQ 1.1 x

107Q. Combining equations and evaluations gives the following

equation for the cell:

L 1 1.37 x 107 (6)

Rv 1 eN u r 1 N u

For example, if the cell were used to measure singly charged

ions (N = 1) having an average mobility u = 1 cm /V/s, and if the
VTVM reading (Figure 2) were Ev = 2 V dc at a bias-or supply
voltage Eb = 400 V dc, Equation (6) indicates that the average
ion or charge carrier population per cubic centimeter of moist
4
air would be Icc = 6.9 x 10
The most precise data taken with the new cell were those
for which the ultrasonic nebulizer was used to humidify the large
box. In a typical experiment starting with a dry box, the humid-
ity was gradually increased by nebulizing a cool water mist or
fog through a plastic tube and into the box. As the humidity
increased, voltage readings (Ev , Figure 2) were taken using the
VTVM at experiments,
in such intervals of 1
it min
was orfound
less.that
As Ehumidification proceeded
v and the equivalent

9
13

1 .W I%
 ion or charge carrier population between the cell plates
[Equation (6)] in the vapor increased very rapidly. At higher
humidities the functional dependence of this increase at a given
temperature was approximately the 13th power of relative
humidity, i.e.,

'cc K1003 (7)

where Kt is the value of Icc at saturation (100% RH) and tended

to increase with temperature, but Kt was by no means constant

between experiments, particularly at higher bias voltages as will
be discussed. The continuity of the bias voltage Eb also
4 affected K t . When bias was applied continuously, Kt was reduced
by a third or more compared with when bias was applied only to
read Ev . Typical values (bias to read) were Kt 2.5 x 10 at
25 *C, and 5.7 x 105 at 35 OC; but regardless of how Kt varied
* for different experimental conditions, the 13th-power dependence
dof
Icc or Ev on RH was remarkably consistent for Eb = 400 V dc
and humidities above 70% RH.

Similar steep humidity dependencies had been observed by

the author for insulator leakage due to surface water films in
earlier experiments carried out to evaluate these leakage
effects. But until the present work, in which the large cell
insulators in the saddle could be dried to very low humidities at
higher temperatures while the cell plates measured moist air
conductivity under much wetter conditions inside the box, it was
not known unequivocally that the ion population between cell
plates in moist air varies as %f(s) 113 , where s = %RH/100, under
typical, humid atmospheric conditions. These results suggest
that the ion or charge carrier species that are present in moist
air and that account for its electrical conductivity with steep
humidity dependence are also present in thin, condensed films on
conductors from which they are released into the moist air in the
presence of an electric field.

Figure 3 shows data for the humidification of the large

box using the ultrasonic nebulizer in an initially dry box, fan
on, Eb = 400 V dc. The data points are closely placed along the
13
line approximating f(s)1 .

14

0%
*1Z
 a!

* 0

0 25"
6d

00

-. 7 08 0:9 .
~S
Figure 3. Ion or Charge Carrier Population per Cubic
Centimeter of Vapor, Icc, vs. Saturation Ratio
(s = %RH/100)

When Eb was raised above 400 V dc in many experiments

(electric field E = Eb/L = 400/0.66 = 600 V/cm), the electrical
properties of moist air became more complex than in smaller
fields where Ohm's law was obeyed. This is illustrated in
Figures 4-6 for bias voltages of 0.4, 1.0, and 1.5 kV, respec-
tively. Note that Figure 4 has a different ordinate scale than
Figures 5 and 6. Temperatures were 24-26 *C. The solid points
show data for humidification of the large box using the ultra-
sonic nebulizer, followed by drying (hollow points) by mixing
with the drier room air without heating. Although the data 13
points cluster about the dashed lines, approximating an f(s)
dependency, the values represented are not always as close to the
equilibrium values as they were, for example, in Figure 3.
Rather, "jumps" or "zigzags" of E v with s were observed. In
effect, the moist air was electrically "noisy." Such behavior

15
 10-6 10

%/
%/

'a 8
107

10- 1-
0
'4
/e

Ion
e/
oor
ChreCrirPpuainoeui
/
o/
/

16
- / Eb=--O.4 kv
Sf
(" /

S
(E b 0. kV

i0-9 10"2 I !
0.7 0.8 0.9 1.0

%_ Figure 4. Data Plots of Voltmeter Reading, Ev (Figure 2),

and Cell Current from Which Average Moist Air
." Ion or Charge Carrier Populations per Cubic
* Centimeter between Cell Plates Can Be Calculated
~(Eb = 0.4 kV)
~16

I-
0o
 10 5 102

106 10 Eb=1 .0 kv

AW
M/

r..
0

'p/ 0
/ 0
7
i-O10 1 0 /

00

/
op
0 - /
,'U /
0/

10"8 10-1 1
0.7 0.8 0.9 1.0
S
Figure 5. Data Plots of Voltmeter Reading, Ev (Figure 2),
and Cell Current from Which Average Moist Air
Ion or Charge Carrier Populations per Cubic
Centimeter between Cell Plates Can Be Calculated
(Eb 1.0 kV)

17

6) C rI
 10-5 102

-S

10-6 10 Eb=1.5 kv 0

/
i 0/
u / 0
a. /o
<

/0

10- 7 1 - /

-S /
/,100 / /

0~//

i10. 8 110.7 I I
0.8 J1
0.9 1.0

S
'Figure 6. Data Plots of voltmeter Reading, E v (Figure 2),

and Cell Current from Which Average Moist Air

Ion or Charge Carrier Populations per Cubic
Centimeter between Cell Plates Can Be Calculated
(Eb = 1.5 kV)

18

%'.~' -
 was observed repeatedly in many experiments; it did not originate
from instrumental artifacts or experimental problems, but was
real. This behavior could be due to changes in the rate at which
ions leave water films on the cell plates and enter the vapor in
response to changing experimental conditions or to sudden changes
in average mobility of the ions corresponding to changes in the
average numbers of water molecules they contained [Equation (6)].
The "jumps" were instantaneous, and the frequently observed
values of Ev had simple integer relationships, e.g., E v might
jump (or drop) by a factor of 2:1, 4:3, 5:2, etc. Examples are
seen in Figure 4. Behavior that could be called "hysteresis-
like" also was observed. An example is shown in Figure 5, where
the (hollow) data points for drying first fell below the dashed
equilibrium line [f(s)13 to the right of this line and then, as
dehumidification continued, crossed over the equilibrium line and
rose above it, where they remained at lower humidities. Such
* hysteresis effects were often seen during humidification as well.
Occasionally, a true "hysteresis loop" w.s formed by data points
about the equilibrium line during cyclical (humidification and
drying) experiments. This behavior first suggested that water
films condensed on the cell plates could be contributing to vapor
. conduction in some previously unknown way.

Yet another interesting property of electrified moist

air seen in Figures 4-6 is that a "knee" or discontinuity occurs
13
where the slope of the curvc changes from f(s) at higher
humidities Lo -f(s) 7 or %f(s) 8 at lower humidities. The location
of this knee is seen to depend on the bias voltage or electric
field strength. In Figure 4, for Eb = 0.4 kV (E = Eb/L =
600 V/cm), the knee is near s = 0.75 (75% RH); in Figure 5, for
Eb = 1.0 kV (E = 1500 V/cm), the knee is ne-ar s = 0.83; in
Figure
bb 6, for Eb = 1.5 kV (E = 2250 V/cm), the knee is near

s = 0.90. From such data plots, it is possible to construct

average or near-equilibrium curves like those shown in Figure 7,
for a variety of bias voltages. Because of hysteresis effects
similar to those seen in Figures 4-6 it was seldom possible to
* reproduce precisely the curves of Figure 7 in experiments; but
these curves are very useful in analyzing the expected electrical
behavior of moist air between the cell plates under given condi-
tions. For example, Figure 7 suggests that above some values of
Eb* perhaps in the range 2 to 3 kV, the ',f(s) 13 dependency of E
b' v
* seen to the right of the dashed line (which is the locus of the
knees seen in Figures 4-6) will no longer be observed regardless
of the saturation ratio, s %RH/100. If so, this could indicate
that above some bias voltage or electric field strength the ions
being measured are multiply charged (N > 1) at all humidities, or
are supplemented by otier kinds of atmospheric ions, and that the
smaller slope of the solid curves to the left of the dashed line
in Figure 7 corresponds to these conditions.

19

W0A
 IVSI - , w wv T t "r Vlw VtT- vrq

10-5 102

Eb

10-6 10 25 0C / 1.5kV
* - / 1.0

/ 0.4

10-7 1 01

W 0.05

~EV

10-10 10-3

10-11 1- 4

0.4 0.6 0.8 1.0

A S
Figure 7. Using Data Like Those in Figures 4-6, These Curves
Were Constructed for Near-Equilibrium Conditions with
Bias Voltage Applied Continuously to the Large Cell

02

02

'dip,
:Z.- From curves like those in Figure 7, one can construct
the average or near-equilibrium curves in Figure 8, which give
important insights into the expected interrelationships between
Ev and E These interrelationships can be investigated experi-
mentally. Figure 8 indicates that for low bias voltages or field
strengths where Ohm's law is obeyed and the ions are singly
charged (to the left of the dashed line), E will increase
linearly with Eb at a given humidity (solid curves) up to some
value of Eb, where Ev will begin to increase more rapidly than
Eb. Experimentally, this was the result found. Figure 9 shows
typical data and indicates that even relatively minor experi-
mental parameters affect the data obtained, such as whether the
bias voltage Eb was increased or decreased as the corresponding
values of E were recorded. In experiments where Ev was measured
v
?.l
over a range of values of Eb, it was preferable to apply the bias
voltage continuously as it was varied, so as to read all values
quickly at a given humidity. This resulted in somewhat lower
* values of Ev being read than if the bias voltage Eb as applied
only to read Ev but the observed functional dependencies were
not otherwise affected.

Very useful experiments were performed in which the

large box was humidified to near-saturation and then dried or
dehumidified while values of Ev versus Eb were recorded at
various humidities. Typical data are shown in Figures 10 and 11.
Figure 10, which depicts humidification of the cell in its box at
23 'C using the ultrasonic nebulizer, shows upturning of the
right-hand ends of the curves for various increasing humidities,
as in Figures 8 and 9. However, as saturation is approached, the
curves begin to flatten. In Figure 11, which is a continuation
of data from Figure 10, the upper curves actually invert and
exhibit flattened tops over the range of bias voltages at the
- maximum humidity reached in these trials (98% RH). This behavior
seems to indicate that when moist air between moist electrodes
0 approaches saturation, some sort of limit or upper value is
reached, at least over a range of bias voltages or field
strengths, where the total population of ions available to carry
all of the electric charges cannot exceed some maximum number.
Figure 12 also shows this curve-flattening, which is seen in
virtually all experiments when "hard" saturation (100% RH) is
achieved by using steam from boiling water to humidify the box.
Here, over the range of Eb shown, the curve-flattening is more
pronounced than in Figure 11 and actually shows negative
resistance characteristics not unlike those exhibited by some
semiconductors; i.e., for the top curves in Figure 12, the cell

21
 ia-5 102

10-7 1

W
M
W C

0.1

icr 9 10-2 A

~ 7 0C
25

1011 10-4
*10-1 10 io310
-. Eb
-. Figure 8. From Figure 7, These Curves Were Constructed for
Near-Equilibrium Conditions with Bias Voltage
Applied Continuously to the Cell

/ 22

0
 -10-

10-6 10

24 0 C

10
10- 0 0
'.5,.

00

-N"

108 101
 10-6 10 23 0C %RH = 97

92
88

84

i~-~ 174

,,' 66

0, > //

.(/ /

10-8 10-1

/
//
/
/
,/
* /

10-9 102 - I I I I
102 103 104
10
Eb

Figure 10. Data Taken for A 4 r at 23 OC Being Humidified

Using Ultrasonic Nebulizer, Fan Off, Continuous
* Bias Voltage Decreasing for Readings

24

7
 .", 10-6 10
98% RH
.., 97
, . 96

" 94

" 10-7 1

%89

A-2. C) Ev

A'.

"..=

10"8 10-1

/23C

1O-9 10-2 _____

% 10 102 10 3
Eb

Figure 11. Continuation of Data in Figure 10 for Air Near

- ,. Saturation, "Drying Down" Without Heating, Pan Off,
Continuous Bias Voltage Decreasing for Readings

0-.
* A 25

d. A-'*"
%
,J
Ji J ~ 'u---- .
*
'
~. '
A ~ ' ~- ~
A-
~
A
~ '
~ V
-.
.' ~
A
'
A
A'A A-A'A
A" A
V * A
'N

,. 10-6 10 -

[j10- 7 1 10

69

, uJ - %RH

; "10.8 101 -270C

J."
1y "Hrd Satraio Sta rooolngWtr

109
0- 10-2 I1 I1
-1 10 102 13
,Eb

' 'Figure 12. Data Taken for Air at 27 *C Being Humidified t

.Q Fan Off, Continuous Bias Voltage

626
 current as measured by E [Equation (2)] actually decreases with
A increasing Eb. This condition is relieved as RH falls toward 96%.
Returning to Figure 11, it can be seen here too that as RH falls,
there is an orderly progression of quite linear curves for
Fv versus E b towards lower values. However, if the "drying down"
curves of Figure 11 are compared with those for initial humidi-
fication in the same experiment (Figure 10), it will be noted
that curves for comparable relative humidities in the two figures
do not show comparable values of Ev for given values of Eb.
Thus, in humidification versus dehumidification, the 'hysteresis"
effects about equilibrium values, as discussed earlier, are again
manifest. This behavior could be explained by the formation of
thin water films on the cell plates during humidification and
delayed evaporation of the films upon drying, provided that the
water films themselves were sources of water ions or charge
carriers.
The following description of moist air conductivity is
suggested by the compilation of all data from the past year's
research.
In the atmospheric electric fields ranging from fair-
weather values (%0.5-100 V/cm or more), moist air conductivity

4 obeys Ohm's law. At a given RH and temperature, the current

between two electrically conductive bcdies is exactly propor-
tional to the field between them.
In this Ohm's law region, increasing RH in a given field
strongly increases the current between the two bodies; the equi-
librium current increase as (RH/100)1 3 . The current increase is
due to surface water films on the bodies, whose thickness
increases with RH, and it arises from charge carriers that leave
the water films and cross the gap between the bodies:

H20 n e H20) n (8)

The number n is of the order 10-12 (Chaimers, 1967). The

current resembles that due to any surface leakage of insulators
supporting the conductive bodies and can be easily mistaken for
such.
As long as the RH is 50-60% or less, the number of
charge carriers is not large compared with other atmospheric
ions. If the bodies are electrodes in a vapor conductivity cell,
their size and separation will determine the observed cell
current, as would be expected for an air-dielectric capacitor.
* But as RH increases, the charge carriers [Equation (8)1 soon
outnumber other ions. If the surface water films are

27

%
"perfect"*, this behavior continues to saturation vapor pressure

where currents up to 105 times larger than those in dry air can
be measured. However, if the surface films on the electrodes are
disturbed under these conditions, e.g., if they coalesce into
bulk liquid water, the cell current can fall precipitously by
orders of magnitude. Measurements show that bulk liquid water
does not produce ions from its surface even in large electric
fields.

For larger atmospheric electric fields of several

hundred volts per centimeter or more, including those leading to
thunderstorm activity, moist air conductivity no longer obeys
Ohm's law but instead becomes nonlinear with field strength.
Although RH still contributes to the current between two conduc-
tive bodies in a given field, the current in the nonlinear region
increases as %(RH/100) 7 - 8 , rather than x(RH/100) 1 3 as in the
Ohm's law region. The total current increases rapidly as ions,
including water ions, are formed in these larger fields by
molecular dissociation:

(H20)rtt+n+ 1 H'(H20)m + OH(H20)n (9)

As in the Ohm's law region, perfect water films on electrodes in

the larger fields of the nonlinear region are found to increase
the conduction current markedly, but bulk liquid water on an
electrode gives no current increase; instead, droplets are pro-
duced that are attracted away from the electrode to the nearest
surface of opposite polarity.

To summarize, in fair-weather atmospheric fields at

higher RHs, surface water films produce large conduction currents
between conductive bodies that are due to charge carriers that
leave the surface water films on the bodies and cannot be
measured by conventional means. In larger fields, including
foul-weather ones, ions are produced by molecular fragmentation
of water and other substances, but surface water films also
contribute to conduction. These effects are fundamental to an
understanding of atmospheric electricity because they could have
a significant effect on the earth's global current.

*In practice, a "perfect" water film can be achieved by fresh

condensation on a smooth surface, suggesting that it is finely
structured. Evaporation of standing water in a closed container
can produce such a film on a surface although its thickness and
thus its activity will not be as great as that for fresh
condensation at saturation humidity on the surface when an
active evaporation source (nebulizer, boiler, etc.) is employed.

28
 4. CONCLUSIONS

Some of the experimental results presented in this

report indicate electrical behavior that has not been previously
observed in atmospheric moist air. Careful technique and repli-
cation of experiments have shown unequivocally that the data
presented here are real and do not arise from insulator leakage
effects, New findings in this work can be summarized as follows:
a. In uniform electric fields of a few hundred volts

per centimeter or less, moist atmospheric air between the cell

electrodes (plates) is comparatively well-behaved. Ohm's law is
obeyed, and conductivities measured can be explained by the
presence of singly charged ions or charge carriers comprising an
average of 10-12 water molecules that carry a single charge from
the thin water film on a given plate to the opposite plate
[Equation (8)]. The total current is found to be steeply depen-
dent upon RH, varying approximately as the 13th power of the
saturation ratio", s = %RH/100.
In other words, the charge
13
carrier density between the plates varies as %f(s) which must
indicate an average or "equilibrium" condition for thin water
* films on plates in humid air. Thus, greater film thicknesses and
unusually high charge carrier densities will occur at high
humidities [Equation (7)1.

b. These equilibria involving cell plates and water

films, whose thicknesses increase with RH in moist air, are
tenuous or "fragile." The methods used to humidify, dry, or heat
an air sample, or to apply the field, can cause variations in
equilibrium that will appear as "hysteresis" effects, i.e., data
-? points will sometimes cross over the equilibrium curve or follow
different paths near it for humidification versus dehumidifica-
tion, as the nature of the surface water film on the cell plates
varies with ambient conditions.

c. In uniform electric fields larger than a few hun-

dred volts per centimeter, the behavior of moist air between the
cell plates becomes more complex. Deviations from Ohm's law
occur, suggesting that molecular fragmentation is producing
0 additional ions [Equation (9)]. The nf(s) 1 3 dependency of con-
ductivity on RH is seen to occur only at higher RHs with increas-
ing field strengths, becoming 'f(s) or 8 at lower RHs.

d. In uniform electric fields larger than a few hun-

g dred volts per centimeter, as saturation humidity is approached,
the behavior of moist air between the cell plates suggests that
their surface water films can only produce a finite or equilib-
rium population of molecular water clusters that are available as
charge carriers [Equation (8)]. When this "limit" is reached,
the total current through the vapor can remain constant over a
* range of field strengths (Figure 11) or, at a given RH, even
exhibit negative resistance characteristics (Figure 12).

29

0L
 e. When the electric field strength reaches 1 to
2 kV/cm or more, the conductivity of moist air can no longer Le
explained by equilibrium populations of water cluster charge
carriers leaving the cell plates [Equation (8)]. The formation
of additional water ions must occur by molecular fragmentation
[Equation (9)], due to the energy then available across the gap
between the cell plates. Ionization increases rapidly and
proportionately with further increases in field strength.
f. If ions (charge carriers) can be produced directly
from thin water films on conductive bodies (including fine water
droplets) in an electric field, and the observed total current
between such bodies is directly proportional to the field
strength, this effect could contribute significantly to the

4detectable
earth's global current; yet the charge carriers would not be
using standard measurement techniques.

30

J.
 .1.

LITERATURE CITED
.

1. Volta, A., "Del Modo di Render Sensibilissima la Piu

Debole Electricita Sia Naturale Sia Artificale," Phil. Trans. Roy
Soc. Vol. 72, pp 237-289 (1782).

2. Faraday, M., Electrostatics and Its Applications,

A.D. Moore, Ed., p 402, John Wiley & Sons, New York, NY, 1973.

3. Lenard, P., "Uber die Elecktrizitat der Wasserfalle,"

Ann. Phys. Lpz. Vol. 46, pp 584-636 (1892).

4. Simpson, G.C., "On the Electricity of Rain and Its

1'.
Origin
(1909).
in Thunderstorms," Phil. Trans. A. Vol. 209, pp 379-413

6 5. Nolan, J.J., and Enright, J., "Experiments on the Elec-

trification Produced by Breaking-up Water with Special Reference
to Simpson's Theory of the Electrification of Thunderclouds,"
Proc. R. Dublin Soc. Vol. 17, pp 1-11 (1922).
6. Blanchard, D.C., "The Electrification of the Atmosphere
by Particles from Bubbles in the Sea," Reference 61-9, Woods Hole
. Oceanographic Institution, 1961.

- 7. Carlon, H.R., "Ion Content of Air Humidified by Boiling

Water," J. Appl. Phys. Vol. 51(1), pp 171-173 (1980).

8. Carlon, H.R., "New Measurements of the Ion Content of

Evaporation-Humidified Air," J. Chem. Phys. Vol. 76(11),
pp 5523-5529 (1982).

9. Carlon, H.R., "New Measurements of the Ion Content of

Evaporation-Humidified Air: Comment," J. Chem. Phys.
Vol. 78(3), pp 1622-1624 (1983).

10. Moore, A.D., Ed., Electrostatics and Its Applications,

pp 390-397, John Wiley & Sons, New York, NY, 1973.

11. Chalmers, J.A., Atmospheric Electricity, pp 86-88,

Pergamon Press, London, 1967.

31

OR0 %%'
",?