Smoked food, method for smoking food and apparatus therefor

Abstract

Food for human consumption is smoked by providing smoke, removing one or more polycyclic aromatic hydrocarbons (PAHs) from the smoke and contacting the food with the treated smoke, or smoking the food and then removing the PAHs from the smoked food. PAH removal is selective, to remove the PAHs that contain 4 or more benzene rings, in particular without significantly changing the volatile profile of the smoked food. A selective filter is provided, as is apparatus comprising a smoke generator, a chamber in which to smoke food, and the selective filter, disposed between the smoke generator and the chamber.

Claims

1. A method of preparing a food, comprising: (a) providing a food, (b) filtering smoke to selectively remove one or more polycyclic aromatic hydrocarbons (PAHs) that contain 4 or more benzene rings therefrom, wherein the filtering comprises filtering the smoke such that the removal of the PAHs that contain 4 or more benzene rings is in greater proportion than the removal of naphthalene and/or phenanthrene, and (c) contacting the treated smoke of (b) with the food of (a), thereby providing a smoke food, wherein the method comprises generating smoke from a source comprising an intimate mixture of (i) a filter comprising clinoptilolite, and (ii) combustion material from which smoke for smoking food for human consumption can be generated.

2. The method of claim 1, comprising selectively removing the PAHs benzo(a)pyrene, benz(a)anthracene, dibenz(a,h)anthracene, two of those PAHs or all three.

3. The method of claim 1, wherein the filtering comprises filtering the smoke selectively to remove the PAHs that contain 4 or more benzene rings relative to removal of smoke flavored components from the smoke.

4. The method of claim 1, wherein the filter comprises pores of size no greater than about 1 nm.

5. The method of claim 1, wherein the filter comprises pores of size no greater than 0.7 nm.

6. The method of claim 1, wherein the level of the one or more PAHs is reduced by in excess of 70% relative to unfiltered smoke.

7. The method of claim 1, wherein the food is a foodstuff, a food additive, a food flavoring and/or a food ingredient.

8. A method of preparing a food, comprising: (a) filtering smoke to selectively remove one or more polycyclic aromatic hydrocarbons (PAHs) therefrom, and (b) contacting the filtered smoke of (a) with a food, wherein the filtered smoke is generated from a composition comprising an intimate mixture of (i) -clinoptilolite and (ii) combustion material from which smoke for smoking food can be generated.

9. The method of claim 8, wherein the combustion material comprises wood.

10. The method of claim 1, wherein the combustion material comprises wood chips, wood dust and/or wood shavings.

Description

(1) Specific embodiments of the invention will now be described in the following Examples with reference to the accompanying drawings in which:

(2) FIG. 1 shows a comparison of volatile profiles of oil smoked with no filter, with a first filter of the invention and with a second filter of the invention, as described in Example 7, the results (1=2-furfural, 2=guaiacol, 3=4-methylguaiacol, A=native clinoptilolite, B=heated clinoptilolite, C=no clinoptilolite) indicating the volatile profiles are not significantly changed;

(3) FIG. 2 shows a schematic diagram of the rig used for smoking experiments incorporating filters of the invention;

(4) FIGS. 3 to 7 shows selected results of removal of PAHs from solution using filters of the invention;

(5) FIG. 8 shows a X-ray diffraction comparison of native filter versus filter heated at 270 C.; and

(6) FIG. 9 shows a X-ray diffraction comparison of native filter versus filter that was ammonium chloride treated and then heated at 270 C.

EXAMPLE 1ANALYSIS OF PAH LEVEL IN SMOKE

(7) Smoke was produced, treated and collected in a pressurised apparatus. Untreated smoke was produced by heating over a gas burner a standard conical flask, which comprised an inlet located in a stopper and a hose outlet located on the side of the flask, containing 20 g of chipped oak (report recites 20, 40 and 100 g). Air was pumped into the inlet carrying smoke out from the flask via a hose on the outlet to a condenser formed by second standard conical flask, which comprised an inlet located in a stopper and a hose outlet located on the side of the flask, cooled by an ice bath.

(8) Uncondensed smoke was carried by the air flow under pump pressure from the condenser and split to the inlets of two identical chromatography columns, a first treatment column containing 10 g of granular clinoptilolite (grain size 1-2.5 mm, pore size 0.39-0.54 nm) and a second control column containing 70 g of ceramic anti-bumping granules and 4 g of sand. Loading the columns in this way provided balanced air/smoke flow through each column. Prior to use, the clinoptilolite was activated by baking at 270 C. for about 12 hours.

(9) Air/smoke from the outlet of each column was bubbled in 40 ml of water to produce two samples for analysis. The samples were analysed by gas chromatography-mass spectroscopy (GC-MS) following solid phase extraction (SPE).

(10) SPE was carried out on each sample using Bond Elut CH SPE columns. Prior to loading the samples, the sorbent in the SPE columns was conditioned with 10 ml methanol followed by 10 ml water-methanol (65:35 v/v).

(11) 21.5 ml of methanol was added to each sample and the samples were loaded to the columns. The columns were then washed three times with 10 ml of water followed by 10 ml of water-methanol. After drying the SPE column under around 70 kPa of vacuum for 30 minutes, the PAHs were eluted with 4 ml of cyclohexane to obtain a sample for GC-MS analysis. A minimum of 69% of PAH was recovered by SPE.

(12) Table 1 shows the levels of six PAHs by treating smoke using the method of the invention.

(13) TABLE-US-00001 TABLE 1 PAH Control Level (ng) Treated Level (ng) Benzo(a)pyrene 0.4 0 Benz(a)anthracene 1.6 0.2 Chrysene 1.6 0.2 Benzo(b)fluoranthene 0.2 0 Indeno(1,2,3-cd)pyrene 0.2 0 Dibenz(a,h)anthracene 0.2 0

(14) For the compounds which were not present in the treated smoke but were present in the control they were either completely removed by the treatment or were at a concentration below the detection limit.

(15) Table 2 shows the percentage reduction in levels of eleven PAHs by treating smoke using the method of the invention.

(16) TABLE-US-00002 TABLE 2 PAH Run 1 (% reduction) Run 2 (% reduction) Phenanthrene 81.1 93.4 Anthracene 87.5 95.6 Fluoranthene 80.1 92.1 Pyrene 82.5 92.8 Benz(a)anthracene 87.9 94.0 Chrysene 86.5 92.4 Benzo(b)fluoranthene 100* 100* Benzo(k)fluoranthene 100* 100* Benzo(a)pyrene 100* 100* Indeno(1,2,3-cd)pyrene 100* 100* Dibenz(a,h)anthracene 100* 100* 100* = PAH detected in control but not in treated sample

(17) Accordingly, treated smoke had significantly lower levels of PAHs, and in some cases removal of PAH below the current limit of detection.

EXAMPLE 2LEVEL OF PAH IN SMOKED FOOD SAMPLES

(18) Aliquots of water and of tomato ketchup and tomato juice (commercially available) were bubbled with filtered smoke treated with the clinoptilolite filter as described in Example 1, all for approximately 3 minutes duration.

(19) The aliquots were tested for taste and compared with controls that had been bubbled with unfiltered smoke in parallel for the same duration.

(20) The effect of using filtered smoke was assessed by tasting the aliquots on the tongue and by drinking the water and the juice. The treated smoke yielded samples that had a distinctive smoky flavour, similar in its smoky components though slightly less intense than the flavour achieved using untreated smoke. The untreated smoke yielded samples that imparted a noticeably astringent, alternatively described as acrid or burning, sensation in the mouth, especially around the jowls. This sensation was absent in the samples smoked using treated smoke.

EXAMPLE 3PREPARATION OF FILTER MATERIALS

(21) A filter of the invention was prepared following this protocol, the resultant material being referred to as treated filter or treated clinoptilolite:

(22) For the Preparation of 600 g of Clinoptilolite Filter Material

(23) Dissolve 636 g ammonium chloride in 6 litres de-ionised water. Add 600 g clinoptilolite. Soak for 24 hours. Stir the mix at timely intervals. Strain the mixture and place recovered clinoptilolite in a tray. Dry at 270 degrees C. for 3 hours.

(24) An alternative filter of the invention was prepared following only the heating element of the protocol, i.e. omitting the pre-treatment with ammonium chloride, the resultant material being referred to as heated filter or heated clinoptilolite.

EXAMPLE 4REMOVAL OF PAHS PRIOR TO SMOKING OF OIL

(25) Materials

(26) Hickory wood chips supplied by Ashwood Smoking Chips Ltd were heated for 3 hours at 130 C. to reduce moisture content. Weight loss after 1 hour was 10.8%, 2 hours 12.0% and 3 hours 12.2%.

(27) Clinoptilolite was supplied by R. S. Minerals Ltd, Guisborough, UK, medium grade (particle size: 1.4 mm, sieve mesh No. 14) and pre-conditioned at 265-285 C. for 3 hours in a Lincat double stone base pizza oven. A further sample of clinoptilolite was soaked in an aqueous solution of ammonium chloride (1M) for 24 hours then heated to 265-285 C. for 3 hours. A control sample of clinoptilolite was used as supplied without heating.

(28) Rapeseed Oil, refined and deodorised, Non GM supplied by BFP Wholesale.

(29) Methodology

(30) A stainless steel smoke rig 1 was set up with a filter column 2 attached directly to a smoke furnace 4, as shown schematically in FIG. 2. A silicon rubber hose 5 was attached to the filter column with the outlet 6 entrained through a pan of rapeseed oil 7, volume 2.5 litres). Heat source 8 combusted the wood chips 9, generating smoke 10 above the chips, which flowed in the direction of the arrows driven by air entering via conduit 11 attached to compressor unit 12. Smoked passed through the filter column 2 and via the hose 5 to bubble through the oil, smoking the oil in the process.

(31) Four Experiments were Undertaken

(32) 1) Controlfilter column empty

(33) 2) filter column containing native clinoptilolite (600 g)

(34) 3) filter column containing heated clinoptilolite (600 g)

(35) 4) filter column containing ammonium chloride treated clinoptilolite (600 g).

(36) Steel wool was used at the base of the filter column to hold the clinoptilolite in place. The smoke furnace was charged with hickory wood chips (1 kg) and heat applied to combust the wood and once smoke flow was well established the smoke hose was placed into the rapeseed oil pan and smoke bubbled through the oil.

(37) Each experiment was conducted for 6 hours. At hourly intervals a sample of rapeseed oil (50 ml) was taken from the oil pan for sensory testing and a sample of clinoptilolite (30 g, from experiments 2, 3 and 4) was taken from the filter column for PAH analysis.

(38) At each interval the wood was checked, stirred and if required more wood was added to maintain constant smoke flow. On completion of each experiment the remaining 6 hour smoked rapeseed oil was collected for sensory testing and PAH analysis.

(39) Extraction of Smoked Oil Samples

(40) Ten grams of oil were placed into a round bottomed flask, and 12 g of potassium hydroxide and 100 ml of ethanol were added. The mixture was subjected to an alkaline treatment with potassium hydroxide and ethanol by heating for 1 h (78 C.) under reflux and filtered. After cooling to room temperature the solution was transferred to a 500 ml separating funnel, and 100 ml of water and 100 ml of cyclohexane were added. The funnel was shaken and the layers were allowed to separate. The ethanol/water phase was transferred into a 250 ml separating funnel and shaken with another 50 ml of cyclohexane. The ethanol/water phase was discarded and the cyclohexane phases were combined. The cyclohexane was washed successively with 50 ml (2 times), 50 ml methanol/water (4:1) and 50 ml of water (2 times). The cyclohexane extract was shaken with 50 ml N,N-dimethylformamide/water (9:1) solution. The layer of N,N-dimethylformamide/water solution was transferred into a 250 ml separating funned, 50 ml of 1% NaCl solution were added and PAH were extracted with 75 ml of cyclohexane. The cyclohexane phase was dried over anhydrous sodium sulphate and concentrated by rotary evaporator under reduced pressure (40 C.). The extract was applied to a SPE column previously conditioned with cyclohexane (5 ml). The flask was rinsed with cyclohexane (3 ml), and the PAH were eluted with 6 ml cyclohexane. The collected fraction was evaporated to approximately 1 mL under a light stream of nitrogen. The concentrated extract were transferred to autosampler vials ready for analysis by GC/MS.

(41) The results are shown in the table below.

(42) TABLE-US-00003 PAH content of oil samples Filter Material Reduc- Reduc- PAH Control Native tion Heated tion napthalene 200533 265411 32%* 110099 45% acenaphthylene 9455 11111 18%* 13689 45%* acenaphthene 8197 7870 4% 3206 61% fluorene 24035 21612 10% 7253 70% phenanthrene 48410 43607 10% 15193 69% anthracene 10964 6032 45% 2654 76% fluoranthene 11896 10331 13% 5135 57% pyrene 11052 8249 25% 4170 62% benz[a]anthracene 1514 711 53% 373 75% chrysene 4276 2936 31% 974 77% benzo[b]fluoranthene 696 660 5% 200 71% benzo[k]fluoranthene 256 450 76%* 133 48% benzo[a]pyrene 520 0 100%* 0 100% indeno[1,2,3- 288 771 168%* 94 67% cd]pyrene dibenz[a,h]anthracene 245 606 147%* 0 100% benzo[ghi]perylene 337 710 111%* 173 49% NB* the results for some 2 benzene-ring containing PAHs and for larger PAHs using native (not heated and not treated) filter material are consistent with observations by others of the effect of raw clinoptilolite on crude extracts containing PAH mixtures. Results using treated filter material (not shown) were very similar to those using heated.

(43) The experiment shows that benzo[a]pyrene was completely removed to below the level of detection by clinoptilolite (native and heated).

(44) There is a significant pattern of reduction of PAHs across the spectrum when heated clinoptilolite is used as a smoke filter in the smoking of oil.

(45) The results further confirm that there appears to be an additional affinity between benzo[a]pyrene and clinoptilolite compared to other PAHs. Benzo[a]pyrene is taken up by heated and heated/treated clinoptilolite more efficiently in the dichloromethane experiment below and in this experiment native clinoptilolite and heated clinoptilolite reduced the benzo[a]pyrene level to below the level of detection.

(46) The most toxic PAHs (Groups 1 and 2A), benzo[a]pyrene (Group1), benz[a]anthracene (Group 2A) and dibenz[a,h]anthracene (Group 2A) were either removed to below the level of detection or in the case of benz[a]anthracene reduced by 75% by heated clinoptilolite.

(47) Hence, heated filter of the invention selectively reduced the levels of the PAHs containing 4 or more benzene rings.

EXAMPLE 5REMOVAL OF PAHS FROM PAH-CONTAINING SOLUTION

(48) An experiment was undertaken to measure the selective removal of PAHs by filters of the invention from a solution of dichloromethane.

(49) Method:

(50) Ammonium Chloride Treatment: native clinoptilolite was stirred with 10 its weight of a 1M solution (53.5 g/L) for 18 h, during which time the ammonium chloride solution was refreshed twice.

(51) Dichloromethane containing 200 micrograms per kg of 16 PAH standards (20 ppm cocktail in dichloromethane diluted with solvent 1:100) was stirred at ambient temperature in the presence of control material and heated and treated filters of the invention. After defined time periods, the solution was decanted off, the PAHs extracted and measured by GC-MS on an EZfaast column using single ion monitoring, specific to each of the PAHs.

(52) Materials:

(53) Standards used were napthalene, acenaphthylene, acenaphthene, fluorene, phenanthrene, anthracene, fluoranthene, pyrene, benz[a]anthracene, chrysene, benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[a]pyrene, indeno[1,2,3-cd]pyrene, dibenz(a,h)anthracene, benzo[ghi]perylene. Nonadecane was used an internal standard.

(54) Clinoptilolite (medium grade sieve mesh No 14) used was either a) native, b) heated to 270 C. in a thin layer for 1 h or c) treated with ammonium chloride and heated to 270 C. in a thin layer for 1 h and d) was a control without clinoptilolite.

(55) Results

(56) The results are in the table below. FIGS. 3-7 show graphically the results for naphthalene, phenanthrene and three relevant known carcinogens benzo(a)pyrene, benz(a)anthracene and dibenz(a.h)anthracene.

(57) TABLE-US-00004 TABLE Results (peak areas) of Control, Native (MU), Heated (MH) and NHCl Treated (MT) Clinoptilolite in a 20 ppb PAH spiked solution of dichloromethane with samples taken at intervals of 1, 5 and 60 minutes. Control MU1 MU5 MU60 MH1 MH5 MH60 MT1 MT5 MT60 Internal standard 152212 183276 178232 171388 154444 171231 161042 153190 155611 159067 (nonadecane) Naphthalene 23040 27368 26575 25712 23063 25461 23374 22098 22754 23282 acenaphthylene 17523 23110 22327 21460 17327 16368 7590 15907 14016 6929 acenaphthene 12599 16685 16243 15219 12361 12351 8387 11813 10648 7050 fluorene 13030 17296 16610 15936 13829 14935 13162 13244 13692 13549 phenanthrene 17854 22566 21542 20543 17696 19186 17389 16524 17265 17090 anthracene 15346 19970 19282 18347 14113 13376 8335 13210 12035 7730 fluoranthene 14616 19333 19053 18158 15088 14742 12673 13935 14582 14188 pyrene 14700 18935 18578 17932 12990 11185 6299 11715 10408 5602 benz[a]anthracene 5158 5560 5038 4243 2365 1670 1072 1425 1380 982 chrysene 7197 8636 8207 7430 4808 3723 2473 3126 3361 2253 benzo[b]fluoranthene 1836 1398 1166 1018 663 583 442 447 467 376 benzo[k]fluoranthene 3873 3802 3368 2845 1347 974 474 620 715 507 benzo[a]pyrene 1430 1326 1127 970 325 125 39 150 74 5 indeno[1,2,3-cd]pyrene 323 240 222 210 159 108 56 100 71 39 dibenz[a,h]anthracene 354 249 214 206 149 112 93 90 79 71 benzo[ghi]perylene 515 428 325 371 181 93 28 93 78 11

(58) The data shows that the larger PAHs are being taken up by the pore structure of the heated and the heated/treated clinoptilolite when in solution resulting in their selective removal from solution.

EXAMPLE 6REMOVAL OF PAHS FROM SMOKED OIL

(59) Oil derived from coconut oil (deodorised, also referred to as MCT oilmedium chain triglyceride oil) was smoked for 72 hours using smoke from oak chips. The resultant oil was stirred with heated filter material for 18 hours and the PAH content of the oil tested and compared with the PAH content of similarly smoked oil not contacted with the filter. PAHs were extracted using saponification and SPE and analysed by GC-MS. Results (expressed as peak areas) are in the table below.

(60) TABLE-US-00005 Control Stirred with heated % PAH PAH (no filter) filter material reduction napthalene 90246 59291 34 acenaphthylene 3514 2591 26 acenaphthene 5297 3080 42 fluorene 16137 7384 54 phenanthrene 35921 30036 16 anthracene 4403 3611 18 fluoranthene 5674 3738 34 pyrene 6908 4444 36 benz[a]anthracene 907 529 42 chrysene 2115 1516 28 benz[b]fluoranthene 651 95 85 benzo[k]fluoranthene 268 128 52 benzo[a]pyrene 347 68 80 dibenz[a,h]anthracene 169 86 49 benzo[ghi]perylene 330 124 62

EXAMPLE 7REMOVAL OF PAHS FROM SMOKED OIL

(61) Rapeseed oil (deodorised) was smoked on the rig using native or heated clinoptilolite to filter out PAHs, with a control smoking using an empty filter bed. Native and heated filter were then compared to empty filter. PAHs were extracted from the oil using saponification and SPE and analysed by GC-MS. Results (expressed as peak areas) are in the table below.

(62) TABLE-US-00006 % PAH % PAH Reduc- Reduc- PAH Empty Native tion Heated tion napthalene 218368 285181 31 118213 46 acenaphthylene 9065 11446 26 13483 49 acenaphthene 8045 8492 6 3523 56 fluorene 23266 21464 8 7012 70 phenanthrene 48328 43595 10 15309 68 anthracene 10282 5622 45 2443 76 fluoranthene 12678 10298 19 5122 60 pyrene 10936 8401 23 3921 64 benz[a]anthracene 1581 645 59 283 82 chrysene 3762 2587 31 722 81 benz[b]fluoranthene 595 565 5 171 71 benzo[k]fluoranthene 443 383 14 126 72 benzo[a]pyrene 731 466 36 112 85 indeno[1,2,3-cd]pyrene 497 488 2 143 71 dibenz[a,h]anthracene 388 333 14 86 78 benzo[ghi]perylene 550 537 2 205 63

(63) Again, selective removal of the 4 or more benzene ring containing PAHs was achieved. These oils were subjected to volatile profile analysissee results presented in FIG. 1. These indicated that there was substantially no change in the volatile profile, with corresponding peak heights and peak areas seen through all traces, hence that the smokey flavour was unchanged.

EXAMPLE 8X-RAY DIFFRACTION ANALYSIS OF FILTER MATERIAL

(64) All X-ray diffraction work was carried out on a Bruker AXS D8 Advance diffractometer operating in transmission capillary mode, equipped with a 2 kW copper anode X-ray tube operating at 40 kV/40 mA and a focussing Ge(111) monochromator to give an incident Cu K1 X-ray beam, =1.54056 . The X-ray detector was a Bruker Lynx Eye position sensitive detector.

(65) Clinoptilolite (Native and Heated)

(66) A sample of clinoptilolite was loaded into a 0.7 mm borosilicate glass capillary and aligned on a goniometer head before mounting on the diffractometer for final alignment. The mounted capillary was rotated during data collection in order to minimise any preferred orientation present. An Oxford Cryosystems Cryostream Compact, mounted co-axially with the sample, allowed temperature control of the sample in the range room temperature to 220 C.

(67) The following general scheme was used for all data collections: Data collection range: 4-45 2 Step size: 0.017 2 Count time per step: 0.6 seconds

(68) The temperature of the Cryostream was ramped from 20 C. to 220 C. in 10 C. increments, then cooled to 20 C., with the sample allowed to equilibrate at each temperature for 5 minutes before a diffraction data collection was started.

(69) Results

(70) The initial scan at 20 degrees C. was compared with the 220 degrees C. scan. There were significant differences, particularly in respect of intensities at low angle between the twoshowing a structural change as a result of water removal by heating.

(71) When datasets at all temperatures were overlaid it was seen that whilst the total changes are significant, there are no major transitions at any given temperature. Rather, the changes appeared continuous.

(72) After standing for hours, the sample appeared much as it did at 220 C. or at 20 C. post-heatinghence there was little change in that time. After standing for days, however, the heated sampled reverted to look much closer to its original pre-heating state.

(73) Separate filter material was heated to 270 C. for about 6 hours. FIG. 8 shows a comparison of native filter versus this heated filter. There are significant changes in the clinoptilolite structure heated at 270 C. compared to unheated control sample.

(74) Thermal Work

(75) A thermogravimetric measurement was carried out on a TA Instruments Q50 thermogravimetric analyser. 53.55 mg of clinoptilolite was accurately weighed into a sample pan and placed in the instrument. The sample was first equilibrated at 30 C. then raised to 270 C. at a rate of 5 C./min, then held at 270 C. to give a total experiment duration of 12 hours. The weight of the sample was recorded over the course of the experiment.

(76) Results

(77) The results showed that the sample undergoes a water loss of ca. 11.5%, with the majority of the water loss occurring by the time the sample reaches 270 C. Water loss is essentially complete after 2 hours.

(78) Ammonium Chloride Treated Clinoptilolite

(79) These data were collected on a Bruker D8 Advance Powder Diffractometer Wavelength=Cu K alpha 1 Geometry=Flat plate reflection Sample presentation=Standard Bruker flat plate sample holder containing powder Scan range=4 to 45 degrees two theta Step size=0.017 degrees Count time=5 s/step Run time per sample=ca. 3.5 hrs

(80) Additional sample preparation info: The samples as supplied were in the form of large granules that were extremely hard and in fact too hard to grind easily using a mortar pestle. As such, each sample was ball-milled in a 5 ml container containing two stainless steel balls for 2 minutes at 30 Hz. This reduced the particle size sufficiently for the samples to be able to be packed into the flat plate sample holder.

(81) Results

(82) Treatment of clinoptilolite with a solution of ammonium chloride followed by heating to 270 C. for 1 hour resulted in significant structural changes in the clinoptilolite matrix, as evidenced by the X-ray diffraction datashown in FIG. 9 (versus native).

(83) There was a marked difference between the native clinoptilolite and the treated clinoptilolite and although the ammonium chloride treated samples showed some of the characteristics of the purely heat treated clinoptilolite samples there were additional peak shifts in other sections of the X-ray profile especially around 22 deg two theta and 30 deg two theta. Hence, the ammonium treated filter exhibited clear and detectable structural changes.

(84) An ammonium chloride treated sample of clinoptilolite added to water then dried at 90 C. showed no tendency to take up moisture and revert back to its original state. Hence, some of those changes were not reversed by water exposure.

(85) Conclusions from X-Ray Work

(86) Drying (achieved in this case by heating) produced significant structural changes, which were continuous with no sudden phase changes. A temperature of 270 C. produced a complete loss of water in 6 hours. The loss of free and bound water accounted for the observed changes. Removal of water improved the activity of the clinoptilolite as a PAH filter. Dried clinoptilolite reverted to its original state in 19 hrs stored at ambient temperature in an open vessel. Dried clinoptilolite did not take up water quickly when stored in a sealed environment or kept in an open capillary. Treatment of clinoptilolite with ammonium chloride produced further changes in the structure compared with heating, changes that were irreversible, such that subsequent uptake of water was prevented or inhibitedenhancing the lifetime and improving the PAH filter efficiency.

(87) Accordingly, the invention provides a method for producing smoke for smoking foods, apparatus therefor, and food produced by these methods.