- Email: [email protected]
Saturated photodissociation of CsI
Volume 50, number
SATURATED
CHKMICAL PHYSICS LETTERS
1
1.5 August 1977
OF Csl *
P~~oTO~~S~~rA~O~
L.W. GROSSMAN +, G.S. HURST, M.C. PAYNE and S.L. ALLMAN Ilealth Physics Divisiotl. Oak Ridge Natimul Laboratory,
Oak Ridge, Temessee 37830, USA
l-very Csl moIeculc in it small volume was photodissociated with B laser pulse; a second pulsed laser detected each C\ atom throu&b re~onanec ioniration spectroscopy. Besides proving one-nrolecule detection, we obtained cross sections for pilotudj~~oci~tjon of 01 2s it function of wzwelength.
1.
Introduction
We report another part of a sequfnce of studies which started with the development of resonance ionization spectroscopy (RIS) [1,2]. In the initial KIS studies it was shown that cncll long-lived XIc(2 1 S) state could be ptlotoionlzc~I in a two-photon process
in which each He(2 *S) was first promored to fIe(3 1 I?) with a 5015-W photon and subsequently ionized with another photon of the same wavelength. In this way the absolute population of the excited He(2 1S) was measured as a function of time followirlg excitation with a proton pulse. More recently [3] it was shown that the saturated two-photon RIS process could be applied
to ato~ns in their grounrl
states.
Thus,
Cs
was
ionited to saturation by using photons at 4555 a to first excite to Cs(7 21BXj2)from which photons at 4555 12 could then ionize the excited atoms. By directing an intense laser beam through a proportional counter, each atom of Cs in a defined volume at the arbitrary time of the Iaser putr;e could be detected. Subsequently, one-atom detection has been used to study density fluctuation [4]. In the present work we show that molecular (e.g., Csl) phot(ldissociation can be used to produce a spak Kcscnrch sponsored by the Energy Research and Develop mcnt Administration under contract with the Union Carbide Corporation. ; Out\ide-the-service: trainee, US Public Health Service, Bureau of tidiolngical i-kalth, Food and Drug Administration,
HIZW.
70
tiatly defined population of atoms in their ground states at a welt-defined time. Such sources of atoms can be combined with time- and space-resolved oneatom detectors in a number’ of novel experimental arrangements. In the course of the prcscnt studies we showed that all of the CsC moIecules in the central portion of a laser beam could be dissociated to neutral Cs and I atoms in their ground states. This, along with the previous demonstrations that each Cs atom can be ionized with another pulsed laser, proves that sin& n~oIecuIes of Csl can be detected. We also obtained new information On the photodissociation of CsT, including the absolute cross section as a function of wavelength.
2. Experiment The experimental arrangement (fig. 1) utilizes a paratlel plate ionization chamber mounted inside a vacuum system. Even when the CsI sample is heated up to 700 K, the concentration of Csl m0ICcules between the parallel plates is quite low; nevertheless, it is more than adcquatc for our method of detection A pulsed laser 2 using kiton red dye pumped with a linear flashlamp is used to dissociate Csl. The narrow beam is about 0.5 mm in diameter when focused and is ideal for defining the initial location of fret Cs For the dissociation of CM molecules we used a linear flashlamppulscd dye laser, model CMX-4, Chrornatix Corpora-
tion, Mountain View, California 94043, USA.
15 August 1977
CHEMICAL PHYSICS LETJXKS
Volume 50, number 1
source laser was set at 3 175 A near the Csl dissociation peak [S]. To dctcct the neutral Cs, the detector laser was set to produce photons at 4593 A which promote CS(G 2S1,2) to Cs(7 2P1j2); and, as appropriate for two-photon RIS, a second photon of the same wavclength photoionized Cs(7 2P,,,). By carrying out the experiment in a buffer gas su& as argon at moderate
pressure (e.g., 100 tom), several desirable effects were produced. First, the CsI molecules were in thermal equilibrium near 300 K due to argon collisions. Also, photoioniration
occurred
from all the degenerate
mag-
nctlc levels. 9inally, wilb a buffer gas, CL-atom diffusion was of no conscqucnce during the lifetime (fwhrn) Fig. 1. I~upcrimcntal arrangement for the study of saturated photodissociatwn of alkah-halide mc;lccules. The pulsed UV laser 1s used to dissociate CsI molcculcs at time t = 0, and the pulsed visible laser ic used to detect Cs atoms at f > 0.
atoms
at time t = 0. A second
pulsed
laser * having
a
of the laser pulse, i.e., 1.5 ps for the source laser and 0.5 ps for the detector laser. For these reasons, most of the studies were made with a buffer gas. Ionization signals were studied as a function of the detector laser fhmice, $bD. We found that as $I~, increased, the ioni?ation signal climbed gradually to a saturated value, con-
much larger beam, about 7 mm in cliametcr, is coaxial with the first narrow beam. In this way the second pulsed laser can bc used to detect those free atoms
sistcnt with earlier work [3] (see fig. 2).
which were libcratcd on the axis of the detector beam at time c = 0 and which are still contained in the dctector cylinder at the arbitrary time t > 0. It will be shown below that the photon flucnce (4~~) associated with the source beam is large enough to dissociate all of the CM n~olcculcs contained in the volume swept
were then made of the photodissociation process. Scvera1 significant observations were maclc on the production of neutrals. For instance, there was a gradual decline in the number of free atoms as the tune between the source laser and detector iascr was increased; but
by the source beam, and the photon fluence (@,) associated with the detector beam is large enough to rcmove one electron from tacit of tllc liberated atoms in the detector volume. The electronics logic and data acquisition are quite conventional and will not be discussed. Each pulse of the detector laser is fired manually when a predctermined condition is met - namely, that it is preceded with a pulse from the source laser at a prescribed time. For each event, three quantities arc recorded: (1) the relative energy per pulse of the source laser, (2) the relative energy per pulse of the detector laser, and (3) a signal proportional to the amount of ionization (i.e., number of free electrons) produced in the guarded region of the parallel plate ionization chamber. In the initial experiments the wavelength of the
With sufficient flucncc Gu to saturate the ionizition, i.e., to detect each free atom in the laser beam, studlcs
this variation was just what one expected on the basis of calculations of the rate of diffusion of the atoms out of the detection region. Special testswere made with very short time delays between the source and detector ,q,
__
p I2 I .e
FRtSH
-----
-
-
DYt
I
I
I
L__
-
A,---___f-
__* ENER6Y
* For the detection of CI atoms WCused a cowial lamp dye laser, model 21OOC, Phnsc-R Co., New Durham, New Hampshire 03855, USA.
---
DENSITY
25 torr ARGON
____*___-I
10
~,ouler/cm2J
Fig. 2. hicasured signals due to Cs atoms as a function of the cncrgy density of the detector laser and at a fixed cncrgy density for the UV laser used to diw~ciatc CII. 71
Volume 50, number
iasers; all of the atoms were dissociated in less than 0.5 ps after the source laser pu!se. No ionization signals due to the source laser alone were observed. Berkowitz f6] observed ionization with a photoionidation mass spcctrometcr only when the photon energy exceeded 7 eV. Fig. 3 shows the ionilration signal as a function of the number of photons in a single pulse of the source laser, both for an unfocused beam and for a beam which was focused with a SO-cm focal length lens. The focused beam signals continue to rise gradually because of a n~~nuniforrn source beam. To Obtain the Cs photoproduction cross section, (T, the following anrdysis was made. Assuming a gaussian beam profile, we write GG) = #o cx~(-_p~/R ‘) 2
(1)
where e. is the flucncc when the radius p = 0, and Ii is a constant. Since each atom is detected, the measured signal is proportional to 00
where tzl is the number of atoms dissociated of length
and N k number
per unit of the Csi molecules.
density
It can be shown that
,O~_.____
-__-
--
--
-
----
--
1
i ig. 3. Signals dtic to Cs atom3 a5 a function of the number of photons in a single pulse of the source I:wx. Each atom produced war, detected with the RIS pracesS, Data WC shown for unfocused and focuwzd hums. ‘Ibc function F’(ago), cq. /3), is fitted
to the c~perimcnt:d
data ~lnd il; ti~c
through the focuted data points.
72
15 August 1977
CHEMICAL PHYSKS LkTTERS
1
curve
v
-
O.‘*L;L_t._=--1 3ooO
SO0
. -..-L.--
32i30 WAVELENGfH
r-t____L 3300
-i.._ 3400
-I
3500
(%I
i?g. 4. Cross section for the p~o~o~roduction of Cs from Csi as a function of wavetength.
where E1 is the exponential integral and r is Euler’s constant (0.577...). The ratio of the focused beam signal to the unfocused beam signal (see fig. 3) is just ~(o~o)~~~~, since in the limit atpo --t 0, F(o&, j = a# For a given total number of photons (e.g., 2.5 X 10 18’), we find the uQlo which makes F(u&,)/u@~ equal to the experiInenta1 ratio (0.41). The tluence Go was determined experimentally by measuring the enera transmitted through a small aperture with a joule meter. In this way we found at 3175 A a vafuc of 2.9 X IO-- i7 cm2 for the Cs photoproduction cross section. Fig. 4 shows the cross section for the production of Cs neutral from CsI as a function of wavelength. The present results have a functional form which is similar to that for photoabsorption [S], and our cross sections at the peak agree to within a few percent with those of ref. [S]. The present measurements were made at 320 K, white the photoabsorpt~~n data were taken at about 1000 K; the reduction in vibrational populations which result from this temperature cIiffcrence could account for the difference in widths. A knowledge of the vapor pressure of CsI was not required to obtain our photodissociation cross sections. In the present technique, the number density of CsI molecults is obtair~cd directly from the saturated curve obtained by plotting the number of Cs atoms per pulse against the laser energy per pulse, or it is obtained through fitting to eq. (3). 3. Discussion
drawn
We have shown that a11Csf molecules in a vohune
Volume
50. nunbcr
1
CHEMICAL
I’IIYSICS LETTt;RS
can bc dissociated by using a photon fluence exceeding 3 X 10” cm-? In - a single laser pulse of microsecond duration. When this fact is combined with the dcmonstratcd fact that one atom can be detected when using an KIS schernc and a proportional counter [3], the capability for one-molecule detection is obvious. Some clarifications of the photodissociation process in Cst cau be deduced from the present studies. Berry [7] showed that in the alkali halide molecules, one may expect to observe in some cases a band system and in other cases a smooth continuum in the photoabsorption spectrum. Here we observe a smooth continuum in the actual appearance of neutral atoms: furthermore, thcsc neutrals appear in a short time, i-c., in less than 0.5 ~.tsafter excitation. Presumably, tRen, WCarc observing a simple process in which photoabsorption occurs from an ionic ground state-to a nonionic predissociation state, which dissociates into the neutral continuum state in a short time, perhaps much less than our measured upper limit of 0.5 ~.ts. Bccausc the prcdissociation state changes from a non-
ionic to an ionic state at a rather large internuclear distance (about 20 K), nuclear motion can no longer be considered adiabatic with respect to electronic motion; hcncc an othcrwisc valid noncrossing rule is violatcd. Our direct observations of prompt neutral atoms
produced in a smooth photon energy continuum, as well as our observation that no Cs+ or I- ions are formed, are consistent with the simple and vivid picture of the alkali photodissociation process painted by Berry [7].
AcknowIedgement WCthank M.H. Nayfeb and J.P. Young for their intcrcst in and support of this work.
References [ 1 J G.S. Ilurst, M.G. Payne, hl.II. Nayfell. J-1’. Jutlisll anti IX. Wajincr, Pflys. Rev. Ixttcrs 35 (1075) 82. [Z] KC. Payne, G.S. Ifurst, hl.11. N.~yfch, I.P. fudish, C.H. Clien, L.B. Wagner md J.1’. Young, Iby\. Rev. Lcttcrs 35 (1975) 1154. 131 G.S. Hurst, h1.H. Nayfch and J.P. Young, hppl. 1’11~4. Letter\ 30 (1977) 229. (4 1 G.S. tiurst, hl.H. Nayfcti and J.P. Young, I’hys. Rev. A, to be publichcd. [S] P. Dnvidovits and D.C. kodhcnd, J. Cl~cm. Plryc. 46 (1967) 2968. [6 ] J. l3crkowitr, J. C’lum. l’l~ys. 50 (1969) 3503. 17 1 R.S. ILrry, J. Chcm. l’hys. 27 (1957) 1288.
73
SATURATED
CHKMICAL PHYSICS LETTERS
1
1.5 August 1977
OF Csl *
P~~oTO~~S~~rA~O~
L.W. GROSSMAN +, G.S. HURST, M.C. PAYNE and S.L. ALLMAN Ilealth Physics Divisiotl. Oak Ridge Natimul Laboratory,
Oak Ridge, Temessee 37830, USA
l-very Csl moIeculc in it small volume was photodissociated with B laser pulse; a second pulsed laser detected each C\ atom throu&b re~onanec ioniration spectroscopy. Besides proving one-nrolecule detection, we obtained cross sections for pilotudj~~oci~tjon of 01 2s it function of wzwelength.
1.
Introduction
We report another part of a sequfnce of studies which started with the development of resonance ionization spectroscopy (RIS) [1,2]. In the initial KIS studies it was shown that cncll long-lived XIc(2 1 S) state could be ptlotoionlzc~I in a two-photon process
in which each He(2 *S) was first promored to fIe(3 1 I?) with a 5015-W photon and subsequently ionized with another photon of the same wavelength. In this way the absolute population of the excited He(2 1S) was measured as a function of time followirlg excitation with a proton pulse. More recently [3] it was shown that the saturated two-photon RIS process could be applied
to ato~ns in their grounrl
states.
Thus,
Cs
was
ionited to saturation by using photons at 4555 a to first excite to Cs(7 21BXj2)from which photons at 4555 12 could then ionize the excited atoms. By directing an intense laser beam through a proportional counter, each atom of Cs in a defined volume at the arbitrary time of the Iaser putr;e could be detected. Subsequently, one-atom detection has been used to study density fluctuation [4]. In the present work we show that molecular (e.g., Csl) phot(ldissociation can be used to produce a spak Kcscnrch sponsored by the Energy Research and Develop mcnt Administration under contract with the Union Carbide Corporation. ; Out\ide-the-service: trainee, US Public Health Service, Bureau of tidiolngical i-kalth, Food and Drug Administration,
HIZW.
70
tiatly defined population of atoms in their ground states at a welt-defined time. Such sources of atoms can be combined with time- and space-resolved oneatom detectors in a number’ of novel experimental arrangements. In the course of the prcscnt studies we showed that all of the CsC moIecules in the central portion of a laser beam could be dissociated to neutral Cs and I atoms in their ground states. This, along with the previous demonstrations that each Cs atom can be ionized with another pulsed laser, proves that sin& n~oIecuIes of Csl can be detected. We also obtained new information On the photodissociation of CsT, including the absolute cross section as a function of wavelength.
2. Experiment The experimental arrangement (fig. 1) utilizes a paratlel plate ionization chamber mounted inside a vacuum system. Even when the CsI sample is heated up to 700 K, the concentration of Csl m0ICcules between the parallel plates is quite low; nevertheless, it is more than adcquatc for our method of detection A pulsed laser 2 using kiton red dye pumped with a linear flashlamp is used to dissociate Csl. The narrow beam is about 0.5 mm in diameter when focused and is ideal for defining the initial location of fret Cs For the dissociation of CM molecules we used a linear flashlamppulscd dye laser, model CMX-4, Chrornatix Corpora-
tion, Mountain View, California 94043, USA.
15 August 1977
CHEMICAL PHYSICS LETJXKS
Volume 50, number 1
source laser was set at 3 175 A near the Csl dissociation peak [S]. To dctcct the neutral Cs, the detector laser was set to produce photons at 4593 A which promote CS(G 2S1,2) to Cs(7 2P1j2); and, as appropriate for two-photon RIS, a second photon of the same wavclength photoionized Cs(7 2P,,,). By carrying out the experiment in a buffer gas su& as argon at moderate
pressure (e.g., 100 tom), several desirable effects were produced. First, the CsI molecules were in thermal equilibrium near 300 K due to argon collisions. Also, photoioniration
occurred
from all the degenerate
mag-
nctlc levels. 9inally, wilb a buffer gas, CL-atom diffusion was of no conscqucnce during the lifetime (fwhrn) Fig. 1. I~upcrimcntal arrangement for the study of saturated photodissociatwn of alkah-halide mc;lccules. The pulsed UV laser 1s used to dissociate CsI molcculcs at time t = 0, and the pulsed visible laser ic used to detect Cs atoms at f > 0.
atoms
at time t = 0. A second
pulsed
laser * having
a
of the laser pulse, i.e., 1.5 ps for the source laser and 0.5 ps for the detector laser. For these reasons, most of the studies were made with a buffer gas. Ionization signals were studied as a function of the detector laser fhmice, $bD. We found that as $I~, increased, the ioni?ation signal climbed gradually to a saturated value, con-
much larger beam, about 7 mm in cliametcr, is coaxial with the first narrow beam. In this way the second pulsed laser can bc used to detect those free atoms
sistcnt with earlier work [3] (see fig. 2).
which were libcratcd on the axis of the detector beam at time c = 0 and which are still contained in the dctector cylinder at the arbitrary time t > 0. It will be shown below that the photon flucnce (4~~) associated with the source beam is large enough to dissociate all of the CM n~olcculcs contained in the volume swept
were then made of the photodissociation process. Scvera1 significant observations were maclc on the production of neutrals. For instance, there was a gradual decline in the number of free atoms as the tune between the source laser and detector iascr was increased; but
by the source beam, and the photon fluence (@,) associated with the detector beam is large enough to rcmove one electron from tacit of tllc liberated atoms in the detector volume. The electronics logic and data acquisition are quite conventional and will not be discussed. Each pulse of the detector laser is fired manually when a predctermined condition is met - namely, that it is preceded with a pulse from the source laser at a prescribed time. For each event, three quantities arc recorded: (1) the relative energy per pulse of the source laser, (2) the relative energy per pulse of the detector laser, and (3) a signal proportional to the amount of ionization (i.e., number of free electrons) produced in the guarded region of the parallel plate ionization chamber. In the initial experiments the wavelength of the
With sufficient flucncc Gu to saturate the ionizition, i.e., to detect each free atom in the laser beam, studlcs
this variation was just what one expected on the basis of calculations of the rate of diffusion of the atoms out of the detection region. Special testswere made with very short time delays between the source and detector ,q,
__
p I2 I .e
FRtSH
-----
-
-
DYt
I
I
I
L__
-
A,---___f-
__* ENER6Y
* For the detection of CI atoms WCused a cowial lamp dye laser, model 21OOC, Phnsc-R Co., New Durham, New Hampshire 03855, USA.
---
DENSITY
25 torr ARGON
____*___-I
10
~,ouler/cm2J
Fig. 2. hicasured signals due to Cs atoms as a function of the cncrgy density of the detector laser and at a fixed cncrgy density for the UV laser used to diw~ciatc CII. 71
Volume 50, number
iasers; all of the atoms were dissociated in less than 0.5 ps after the source laser pu!se. No ionization signals due to the source laser alone were observed. Berkowitz f6] observed ionization with a photoionidation mass spcctrometcr only when the photon energy exceeded 7 eV. Fig. 3 shows the ionilration signal as a function of the number of photons in a single pulse of the source laser, both for an unfocused beam and for a beam which was focused with a SO-cm focal length lens. The focused beam signals continue to rise gradually because of a n~~nuniforrn source beam. To Obtain the Cs photoproduction cross section, (T, the following anrdysis was made. Assuming a gaussian beam profile, we write GG) = #o cx~(-_p~/R ‘) 2
(1)
where e. is the flucncc when the radius p = 0, and Ii is a constant. Since each atom is detected, the measured signal is proportional to 00
where tzl is the number of atoms dissociated of length
and N k number
per unit of the Csi molecules.
density
It can be shown that
,O~_.____
-__-
--
--
-
----
--
1
i ig. 3. Signals dtic to Cs atom3 a5 a function of the number of photons in a single pulse of the source I:wx. Each atom produced war, detected with the RIS pracesS, Data WC shown for unfocused and focuwzd hums. ‘Ibc function F’(ago), cq. /3), is fitted
to the c~perimcnt:d
data ~lnd il; ti~c
through the focuted data points.
72
15 August 1977
CHEMICAL PHYSKS LkTTERS
1
curve
v
-
O.‘*L;L_t._=--1 3ooO
SO0
. -..-L.--
32i30 WAVELENGfH
r-t____L 3300
-i.._ 3400
-I
3500
(%I
i?g. 4. Cross section for the p~o~o~roduction of Cs from Csi as a function of wavetength.
where E1 is the exponential integral and r is Euler’s constant (0.577...). The ratio of the focused beam signal to the unfocused beam signal (see fig. 3) is just ~(o~o)~~~~, since in the limit atpo --t 0, F(o&, j = a# For a given total number of photons (e.g., 2.5 X 10 18’), we find the uQlo which makes F(u&,)/u@~ equal to the experiInenta1 ratio (0.41). The tluence Go was determined experimentally by measuring the enera transmitted through a small aperture with a joule meter. In this way we found at 3175 A a vafuc of 2.9 X IO-- i7 cm2 for the Cs photoproduction cross section. Fig. 4 shows the cross section for the production of Cs neutral from CsI as a function of wavelength. The present results have a functional form which is similar to that for photoabsorption [S], and our cross sections at the peak agree to within a few percent with those of ref. [S]. The present measurements were made at 320 K, white the photoabsorpt~~n data were taken at about 1000 K; the reduction in vibrational populations which result from this temperature cIiffcrence could account for the difference in widths. A knowledge of the vapor pressure of CsI was not required to obtain our photodissociation cross sections. In the present technique, the number density of CsI molecults is obtair~cd directly from the saturated curve obtained by plotting the number of Cs atoms per pulse against the laser energy per pulse, or it is obtained through fitting to eq. (3). 3. Discussion
drawn
We have shown that a11Csf molecules in a vohune
Volume
50. nunbcr
1
CHEMICAL
I’IIYSICS LETTt;RS
can bc dissociated by using a photon fluence exceeding 3 X 10” cm-? In - a single laser pulse of microsecond duration. When this fact is combined with the dcmonstratcd fact that one atom can be detected when using an KIS schernc and a proportional counter [3], the capability for one-molecule detection is obvious. Some clarifications of the photodissociation process in Cst cau be deduced from the present studies. Berry [7] showed that in the alkali halide molecules, one may expect to observe in some cases a band system and in other cases a smooth continuum in the photoabsorption spectrum. Here we observe a smooth continuum in the actual appearance of neutral atoms: furthermore, thcsc neutrals appear in a short time, i-c., in less than 0.5 ~.tsafter excitation. Presumably, tRen, WCarc observing a simple process in which photoabsorption occurs from an ionic ground state-to a nonionic predissociation state, which dissociates into the neutral continuum state in a short time, perhaps much less than our measured upper limit of 0.5 ~.ts. Bccausc the prcdissociation state changes from a non-
ionic to an ionic state at a rather large internuclear distance (about 20 K), nuclear motion can no longer be considered adiabatic with respect to electronic motion; hcncc an othcrwisc valid noncrossing rule is violatcd. Our direct observations of prompt neutral atoms
produced in a smooth photon energy continuum, as well as our observation that no Cs+ or I- ions are formed, are consistent with the simple and vivid picture of the alkali photodissociation process painted by Berry [7].
AcknowIedgement WCthank M.H. Nayfeb and J.P. Young for their intcrcst in and support of this work.
References [ 1 J G.S. Ilurst, M.G. Payne, hl.II. Nayfell. J-1’. Jutlisll anti IX. Wajincr, Pflys. Rev. Ixttcrs 35 (1075) 82. [Z] KC. Payne, G.S. Ifurst, hl.11. N.~yfch, I.P. fudish, C.H. Clien, L.B. Wagner md J.1’. Young, Iby\. Rev. Lcttcrs 35 (1975) 1154. 131 G.S. Hurst, h1.H. Nayfch and J.P. Young, hppl. 1’11~4. Letter\ 30 (1977) 229. (4 1 G.S. tiurst, hl.H. Nayfcti and J.P. Young, I’hys. Rev. A, to be publichcd. [S] P. Dnvidovits and D.C. kodhcnd, J. Cl~cm. Plryc. 46 (1967) 2968. [6 ] J. l3crkowitr, J. C’lum. l’l~ys. 50 (1969) 3503. 17 1 R.S. ILrry, J. Chcm. l’hys. 27 (1957) 1288.
73