1

PARTICULATE NITRATE MEASUREMENTSDURING CRPAQS

Judith C. ChowJohn G. WatsonDouglas H. LowenthalDesert Research Institute, Reno, NV

Susanne V. HeringAerosol Dynamics Inc., Berkeley, CA

presented at the

CRPAQS Data Analysis MeetingSacramento, CA

March 9, 2004

2

OBJECTIVES

• Present nitrate measurements fordifferent samplers, locations, andtime periods

• Evaluate the extent of nitratevolatilization

3

INTEGRATED SAMPLERS

Dual-channelsequential filter

samplerDesert Research

Institute, Reno, NV

RAAS-400PM2.5 speciation sampler

Andersen Instruments, Smyrna, GA

RAAS-100single-channel PM2.5 FRM sampler

Andersen Instruments, Smyrna, GA

Minivol samplerAirmetrics,Eugene, OR

4

Andersensingle-channel FRM

Louvered SA-246PM10 inlet

Frontquartz-fiber

Non-volatilizednitrate IC

16.7 L/min

PM2.5WINS

impactor

Variable-speedpump

Andersen RAAS louvered inlet

Teflon-coated modifiedAIHL a

PM2.5 cyclone inlet

FEP Teflon-coatedmodified AIHL

PM2.5 cyclone inlet

Manifold 2

Frontquartz-fiber

Non-volatilizedcarbon TORNitrate IC

Backupquartz-fiber

Adsorbedgaseous

carbon TOR

Frontquartz-fiber

Non-volatilizednitrate IC

Backupnylon

Volatilizednitrate IC

Manifold 1

24 L/min

Channel 27.3 L/min

to otherchannels

Channel 58.7 L/min 15.3 L/min

nitricacid

denuder

Critical orifice flow controller

24 L/min

fieldblank

no flow

to otherchannel

16.7 L/min

Desert Research Institutesequential filter sampler (SFS)

Bendix 240 cPM2.5 cyclone inlet

to otherchannel andbypass flow

93 L/min

113 L/min

Manifold

20 L/min

Differential pressure flow controller

fieldblank

no flow

Frontquartz-fiber

Non-volatilizednitrate IC

nitricacid

denuder

Backupsodium-chloride-

impregnated cellulose

Volatilizednitrate IC

Airmetrics Minivol sampler

Louvered SA-246PM10 inlet

5 L/min

PM2.5 impactor

PM10 impactor

Front

quartz-fiber Non-volatilized nitrate IC

Backup sodium-chloride-

impregnated cellulose

Volatilized nitrate IC

Diaphragm pump and flow set valve

INTEGRATED SAMPLERS (continued)

5

TYPES OF INLETSCut Point Slope Flow

Type (d50) (√d84/d16) Rate

EPA WINS 2.48 µm 1.18 16.7 L/minimpactor (FRM)

AIHL cyclone 2.5 µm 1.18 24 L/min(RAAS)

Bendix/Sensidyne 2.5 µm 1.7 113 L/min240 cyclone (SFS)

Impactor 2.5 µm 1.23 5 L/min(Minivol)

INDEX BAR (PM2.5B HASSPACERS UNDER THE BARAND LONGE R SCREWS)

6

TYPES OF DENUDERS

SamplingSystem Denuder Type

FRM None

RAAS Sodium-carbonate-coated glass denuder(flow rate = 8.7 L/min)

SFS Aluminum-oxide-coatedtubular denuder(flow rate = 20 L/min)

Minivol None

7

TYPES OF FILTER HOLDERSInlet surface Filter holder

FRM Anodized Delrin ringaluminum with stainless

steel grid

RAAS FEP Teflon Delrin ring,coated polycarbonate

holder

SFS Anodized Nucleporealuminum polycarbonate

holder

Minivol Anodized Nucleporealuminum polycarbonate

holder

8

SAMPLING SUBSTRATES FORION CHROMATOGRAPHY ANALYSIS

Analyte Filter type Sampling system

Non-volatilized Pallflex FRMNO3

– #2500 QAT-UP RAASquartz-fiber SFS

Minivol

Volatilized Schleicher & Schuell RAASNO3

– #66 nylon

Volatilized Whatman 31ET SFSNO3

– cellulose-fiber Minivolimpregnated with NaCl

9

NH4NO3 (p) <=> NH3 (g) + HNO3 (g)

low T, high RH <=> high T, low RH

THE AMMONIUM NITRATE EQUILIBRIUM

10

T (oC)

0 5 10 15 20 25 30 35

RH

(%)

20

40

60

80

100

BakersfieldFresno

24-HOUR AVERAGE TEMPERATURE (T) ANDRELATIVE HUMIDITY (RH) (Dec. 1999–Jan. 2001)

T and RH aresimilar at bothsites and varyinversely.

11

Date

12/1/99 4/1/00 8/1/00 12/1/00

Vola

tiliz

ed N

O3– /T

otal

Par

t. N

O3– (%

)

0

20

40

60

80

100

T (o C

)

5

10

15

20

25

30

35

Angiola BakersfieldFresnoBakersfield Temp.Fresno Temp.

Peaks duringwarmer months.Relatively lessvolatilization atAngiola.

MONTHLY AVERAGE NITRATE VOLATILIZATION

12

Peaks in winter and fall.

Volatilization is highestin spring and summerat the two urban sites.

Angiola

Winter Spring Summer FallN

itrat

e (u

g/m

3 )

0

5

10

15

20Non-VolatilizedVolatilized

Bakersfield

Winter Spring Summer Fall

Nitr

ate

(ug/

m3 )

0

5

10

15

20Non-VolatilizedVolatilized

Fresno

Winter Spring Summer Fall

Nitr

ate

(ug/

m3 )

0

5

10

15

20Non-VolatilizedVolatilized

PM2.5 PARTICULATENITRATE (PNO3

–)

13

Nitratevolatilizationincreasesthroughout theday withtemperature,even in winter.

WINTER INTENSIVE NITRATE VOLATILIZATIONBY SAMPLE PERIOD

CRPAQS Winter IntensiveNO3

– Volatilization by Period

Sampling Period (PST)

0-5 5-10 10-13 13-16 16-24

Vola

tiliz

ed N

O3– /T

otal

Par

t. N

O3– (%

)

0

5

10

15

20

T (o C

)

-2

0

2

4

6

8

10

12

14

16

-2

0

2

4

6

8

10

12

14

16

AngiolaBakersfieldBethel Isl.FresnoSierra NVBakersfield T oCFresno T oC

14

Sampling Date during 200010

-1410

-1610

-1710

-1810

-1910

-2010

-2210

-2310

-24 11-2

11-3

11-4

11-5

11-6

11-7

11-8

11-9

Vola

tiliz

ed N

O3– /T

otal

Par

t. N

O3– (%

)

0

20

40

60

80

100

Fres

no T

(o C)

8

10

12

14

16

18

20

22

AngiolaFresnoFresno Temp.

Degree ofvolatilizationdecreases withtemperatureat both sitesthrough theepisode.

FALL INTENSIVE DAILY NITRATE VOLATILIZATION

15

SFS (higherface velocity)volatilizesmore nitratethan the RAASsampler(lower facevelocity).

Date12

/1/99

1/1

/00

2/1/00

3/1

/00

4/1/00

5/1

/00

6/1/00

7/1

/00

8/1/00

9/1

/00

10/1/

00

11/1/

00

12/1/

00

1/1/01

2/1

/01

Vola

tiliz

ed N

O3– /T

otal

Par

t. N

O3– (%

)

0

20

40

60

80

100

SFS/

RAA

S

0

2

4

6

8

10

12

14

16SFSRAASSFS/RAAS

10.5 cm/s

24.2 cm/s

MONTHLY AVERAGE NITRATE VOLATILIZATION

16

VOLATILIZED NH4NO3NOT MEASURED BYGRAVIMETRIC MASS

Angiola Daily Nitrate Volatilization

Date12/1/99 2/1/00 4/1/00 6/1/00 8/1/00 10/1/00 12/1/00 2/1/01 Vo

latil

ized

NH

4NO

3/PM

2.5 (

%)

0

10

20

30

40

50

60

Bakersfield Daily Nitrate Volatilization

12/1/99 2/1/00 4/1/00 6/1/00 8/1/00 10/1/00 12/1/00 2/1/01 Vola

tiliz

ed N

H4N

O3/P

M2.

5 (%

)

0

10

20

30

40

50

Fresno Daily Nitrate Volatilization

Date12/1/99 2/1/00 4/1/00 6/1/00 8/1/00 10/1/00 12/1/00 2/1/01 Vo

latil

ized

NH

4NO

3/PM

2.5 (

%)

0

20

40

60

80

100

Range = 0-50%Average = 9.5%

Range = 0-46%Average = 15.4%

Range = 0-80%Average = 21%

Date

PM2.5 mass isunderestimated by up to80% on a daily basis dueto nitrate volatilization.

17

VOLATILIZED NH4NO3NOT MEASURED BYGRAVIMETRIC MASS

Nitrate volatilizationexceeded 20% only whenPM2.5 < 20-30 µg/m3.

Angiola

PM2.5 (µg/m3)

0 20 40 60 80 100 120 140 160Vola

tiliz

ed A

MN

IT/P

M2.

5 (%

)

0

20

40

60

80

100

N=51

Bakersfield

PM2.5 (µg/m3)

0 20 40 60 80 100 120 140 160Vola

tiliz

ed A

MN

IT/P

M2.

5 (%

)

0

20

40

60

80

100

N=62

Fresno

PM2.5 (µg/m3)

0 20 40 60 80 100 120 140 160Vola

tiliz

ed A

MN

IT/P

M2.

5 (%

)

0

20

40

60

80

100

N=64

18

In summer,SFS PNO3

– > FRM NO3–

4.8 > 2.4 µg/m3

(average)

In winter,SFS PNO3

– < FRM NO3–

13 < 16 µg/m3

(average)

Fresno

SFS PNO3– (µg/m3)

0 2 4 6 8 10

FRM

Non

-Vol

atiliz

ed N

O3– (u

g/m

3 )

0

2

4

6

8

10

y/x = 0.48±0.32N = 28

Fresno

SFS PNO3– (µg/m3)

0 10 20 30 40 50 60 70

FRM

Non

-Vol

atiliz

ed N

O3– (u

g/m

3 )

0

10

20

30

40

50

60

70

y/x = 1.12±0.25N = 16

Summer

Winter

19

PM2.5 mass and sulfatefor FRM aresystematically higherthan SFS.

This could explain in partwhy winter FRM non-volatilized NO3

– is higherthan SFS total particulateNO3

–.

Fresno sulfate

SFS SO4= (µg/m3)

0 2 4 6 8

FRM

SO

4= (µg/

m3 )

0

2

4

6

8

y/x = 1.11±0.13 N = 57

Fresno mass

SFS PM2.5 (µg/m3)0 20 40 60 80 100 120 140 160

FRM

PM

2.5 (

µg/m

3 )

0

20

40

60

80

100

120

140

160y/x = 1.06±0.25 N = 58

20

SFS particulate (denuded)nitrate is higher thanMinivol (undenuded)nitrate in summer, butsimilar in winter.

Bakersfield

Date

12/1/

991/1

/002/1

/003/1

/004/1

/005/1

/006/1

/007/1

/008/1

/009/1

/0010

/1/00

11/1/

0012

/1/00

1/1/01

2/1/01

Parti

cle

NO

3– (Non

-Vol

atiliz

ed +

Vol

atiliz

ed) (

ug/m

3 )

0.1

1

10

100

1000

SFS (denuded)Minivol (non-denuded)

Sierra NV

Date

12/2/

0012

/8/00

12/14

/0012

/20/00

12/26

/00

1/1/01

1/7/01

1/13/0

11/1

9/01

1/25/0

1

Parti

cle

NO

3– (Non

-Vol

atiliz

ed +

Vol

atiliz

ed) (

ug/m

3 )

0

5

10

15

20

25

SFS (denuded)Minivol (non-denuded)

21

Angiola Winter Intensive

PNO3– (ug/m3)

0 10 20 30 40 50 60R&P

840

0N N

O3– (u

g/m

3 )

0

10

20

30

40

50

60

Bakersfield Winter Intensive

PNO3– (ug/m3)

0 20 40 60 80 100 120R&P

840

0N N

O3– (u

g/m

3 )

0

20

40

60

80

100

120

Fresno Winter Intensive

PNO3– (ug/m3)

0 10 20 30 40 50 60 70R&P

840

0N N

O3– (u

g/m

3 )

0

10

20

30

40

50

60

SFS PNO3– and R&P NO3

–

agree up to 20-30 ug/m3,but R&P NO3

– is lower at

higher concentrations –most strongly at Angiola,least at Bakersfield.

This may be related tothe difference betweenambient and R&Psampling temperature.

22

Continuous andfilter HNO3 bydenuder differenceare only moderatelycorrelated.

CRPAQS Winter Intensive by Period at Fresno

Filter Denuder Difference HNO3 (µg/m3)

0 2 4 6 8 10 12

Con

tinuo

us H

NO

3 (µg

/m3 )

0

2

4

6

8

10

12

r = 0.61

23

Fresno Monthly HNO3 and Particulate Nitrate

Date

11/1/

00

1/1/01

3/1

/01

5/1/01

7/1

/01

9/1/01

11

/1/01

1/1

/02

3/1/02

5/1

/02

Tota

l Par

ticul

ate

NO

3 (µg

/m3 )

0

5

10

15

20

25

HN

O3 (

µg/m

3 )

0

1

2

3

4

5RAAS Total Part. NO3

Continuous HNO3

PNO3- peaks in

winter, while HNO3peaks in summerwhen it is dominant,as expectedthermodynamically.

24

CONCLUSIONS

• The degree of ammonium nitrate volatilization variesseasonally and diurnally with temperature.

• Ammonium nitrate volatilization also increases withsampling face velocity where the SFS (24.2 cm/s) >RAAS (10.5 cm/s).

• Volatilized ammonium nitrate is up to 80%, 50%, and46% of measured daily PM2.5 mass at Fresno, Angiola,and Bakersfield, respectively.

• Nitrate volatilization exceeds 20% only whenPM2.5 < 20–30 µg/m3.

25

CONCLUSIONS (continued)

• Total particulate nitrate measured by denuded SFS agreedwell with un-denuded Minivol sampler non-volatilized andvolatilized nitrate in winter but not in summer at Bakersfield.

• Continuous R&P nitrate agreed with SFS particle nitrate upto 20-30 µg/m3 but was lower at higher concentrations.

• Filter-based nitric acid measured by denuder difference wasmoderately correlated (r=0.61) with continuous nitric acid inFresno.

• Fresno gaseous nitric acid concentrations peaked at about 4µg/m3 in summer coinciding with the annual minimum inparticle nitrate.