Indoor Air Quality in a Photo Archive Without Climate Control.
A Case Story

Karen Brynjolf Pedersen & Morten Ryhl-Svendsen

National Museum of Denmark

The collection and the archive

Fig. 1: The archive room

This collection contains approximately 60,000 photographic negatives and positives, mostly glass plates, but also cellulose acetate and polyester sheet negatives. Until 2000 the archive also contained cellulose nitrate negatives, however these have been removed and stored separately due to the flammable properties of cellulose nitrate.
Photographic techniques within the collection include black-and-white collodium negatives, black-and-white gelatine negatives, autochrome colour plates and modern colour slides. The material dates from ca. 1870s to today.

The photographs are housed in paper and glassine envelopes, stored in drawers in metal cabinets. The archive is located in the Museum's attic. The room is located in a wing with the busy street Stormgade just outside, which is passed by ca. 18,000 vehicles each working day. The room itself is not designed especially for archive purposes. It has no air conditioning or mechanical ventilation, only a draughty window.
The cabinets are sticky and dusty appearance at surfaces, and there is black and "sandy" dust in the window recess.

Recommendations from standards on storage conditions
For photographic materials international standards exist for all aspects of storage: archive climate, housing materials etc. For "Extended-term storage" (for items of permanent value) there are the following recommendations [1]:

ISO 18911:2000:
        Triacetate: 2°C at 20-30% RH
        Polyester: 21°C at 20-50% RH

ISO 18918:2000:
        Glass plates: 18°C at 30-40% RH

In addition, a general recommendation of the ISO standards is to "Avoid abrasive particulates, and harmful air pollutants".

Other standards for paper based archival material have more specific air pollution limitations:

NBSIR 83-2795:
        NO₂: max 5 µg/m³

BS 5454:2000:
        NOx: max 5-10 µg/m³

The survey
Considering the above description of the archive, we investigated the following:

• the condition of items in the collection?
• the climate: level and variation?
• air pollution levels: does the building envelope ensure sufficient protection?
• the dust load: origins and level?

Methods and results
Condition of negatives
Previous records of the collections condition are few. However, in 1988 an unspecified number ("many") of negatives were discarded, because they were bleached out. In 1999 a condition survey covering 61 random sampled plates showed that 51 had silver mirror - a chemical damage to silver images, which can be caused by oxidative air pollutants or bad enclosure materials. In the same survey it was noted that all negative envelopes were dirty.

Climate
The archive is not continuously climate monitored, which is why no previous data is available. We have installed a temperature and relative humidity logger in the room in order to monitor the climate Throughout 2003 the temperature and relative humidity have been monitored in the archive and logged with hourly intervals. The data is shown in fig 2.

Fig. 2: Climate in archive: Temperature (red) ; Relative Humidity (blue)

The temperature varied during 2003 from 11 to 28°C, with a mean value of 20°C. The temperature was within our target value of 18°C +/- 5° for 82% of the year. The relative humidity varied between 27% and 51%, with a mean value of 39%RH. With a target value of 30-40%RH, according to ISO 18918:2000 (glass plates), the climate was within these limits for only 44% of the year 2003 (fig 3).

In other words, the yearly trend is such that the temperature will vary from an average of 15-16° during winter, up to 25-26° in summer. The RH will top at ca. 50% in summer, and fall to 30% during winter and spring. That corresponds to a Time Weighted Preservation Index (TWPI) of approx. 50 years [2]. The preservation index will be highest in winter where the indoor climate is cool and dry, and much lower in summer where the climate is warm and humid (fig. 4).

Air pollution
It was obvious from the dust layer around the window that outside air leaked in. The pollution in outdoor air in central Copenhagen is dominated by traffic exhaust gases and particulates. The outdoor concentration of O₃ and NO₂ is typically in the range of 25-50 µg/m³, however during rush hours, episodes of concentrations up to 150 µg/m³ occur.
From earlier investigations in nearby, similar locations within the museum building, the indoor concentration, and the indoor/outdoor (I/O) rate of the air pollutants O₃, NO₂ and NOx had been determined:

November-December 1998 [3]:
NO2 indoors:	16-19 µg/m3	I/O approx. 0.5
For the one year period 17/9-2001 to 16/9-2002 [4]:
O3 indoors:	3 µg/m3	I/O: 0.05
NO2 indoors:	22 µg/m3	I/O: 0.66
NOx indoors:	36 µg/m3	I/O: 0.81

The concentration of O₃, NO₂ and NOx was now measured in the archive for four months in 2003. The measurements were carried out by passive sampling, by four successive sampling periods of about one month each. The sampler type used was Analyst, as described by De Santis et al [5].
The results were compared with the outdoor pollution levels [6].

The result was surprising: while the I/O ratio for O₃ was quite low, as expected, for three out of the four months the I/O rates for NO₂ and NOx were 1 or higher, indicating that a source of nitrogen oxides was present in the archive, which drowned the contribution from the outdoor environment (fig. 5).
The recommendation for levels of NO₂ is 5 µg/m³. In the archive between 10-50 µg/m³ were measured . The recommendation for levels of NOx is between 5-10 µg/m³. In the archive between 200-300 µg/m³ were measured, except in May where "only" 25 µg/m³ was measured.

Dust
Horizontal surfaces in the archive were generally quite dirty. The appearance of the dirt differed between surfaces in the open room (fluffy dust) and the surfaces inside the drawers for the negatives (black, sticky). Samples of durst was taken using Dust Lifter Gel [7] . The samples were digitally photographed at 40X, and analysed by the image analysis software ImageJ [8] . Analysis of particle size and distribution, particle count, and area covering fraction was carried out. Only coarse particles were investigated, with a diameter above 2 µm.

The area fraction covered with dust in the archive was high:
Inside drawers the covering fraction was approximately 14%, on top of the cabinets approximately 41%.

These layers are "old dust", maybe due to decades of deposition. The dust that fell on a new, clean surface (petri dish) covered 1.03% on top of a cabinet, and just 0.03% inside a drawer, after 74 days.

However, deposited particles in drawers are smaller and much more numerous, than out in the room (Fig. 6). Analysing the "old dust" layer we found, that on 1 mm³ on top of a cabinet, the average number of particles were 2902, with an average size of 130 µm². In a drawer, the average number of particles was 9807, with an average size of just 11 µm².

Fig. 6: Left: Dust layer from top of cabinet. Right: Dust layer from inside drawer

Investigating the dust samples by light microscopy, we identified textile fibres, skin flakes, hair, mineral wool fibres, sand grains and pollen.
By Scanning Electron Microscopy (LW-SEM) also soot, gypsum, calcium carbonate, paint flakes (containing titanium), asbestos and NaCl were identified. In the drawer we mainly found wood fibres and titanium (from paper envelopes).

Fig. 7: SEM images. Left: Dust from top of cabinet. Right: Paper fibre (wood) from inside drawer

Dust samples taken by sweeping a clean filter paper over a surface were analysed for their organic content. Analysis was carried out by GC/MS, after the samples were extracted in methylene chloride by soxhlet method.
We found fatty acids, carboxylic acids (lauric, myristic, palmitic, stearic), alkanes (C₂₀-C₃₀) and pthalates.

Discussion
The temperature and RH is too high in summer. This is clearly illustrated on the TWPI diagram in Fig. 4. This is especially a problem for negatives or slides on plastic supports, and for image dyes in colour photographs. While complete air conditioning is not possible in this archive, dehumidification during summer would be an option. This will lower the rate of chemical decay, and the climate could be not too far off the ISO recommendations for glass plates. Deteriorated plastic sheets, however, should be safety copied, and removed to better (cooler) archive conditions.

If a dehumidifier is to run properly, the room should be made more airtight. The window must therefore be replaced with a more tight construction. This will also reduce dust from entering from outdoors. This refers especially to the silicates, soot, pollen and NaCl found in the dust on the cabinets and in the window recess.
The dirt found inside the drawers are mainly of indoor origin; the carboxylic and fatty acids are likely to be remains of cleaning detergents, human skin flakes etc. The mineral wool fibres origins most probably from a recent repair of the ceiling, as does the gypsum, calcium carbonate, and paint flakes.
The dust load seems high, but it is mainly "old dust", due to decades of absence of cleaning. The amount of dust that fell over 74 days is much lower than we see in places with everyday human activity, e.g. in exhibition galleries.
Photographic materials are very sensitive to air pollutants, as described in ISO 18911:2000 [9]. However we did not find any damages on the negatives that we specificly could relate to the high rate of air pollution or dust. However, the silver mirroring on gelatine glass plate negative is due to oxidation of the image silver by contaminants, whether these come from air or from storage materials.

We recommend that the archive must undergo a major cleaning, and that a regular cleaning routine is introduced thereafter. In order to remove the dirt layer which has accumulated on the old envelopes, those should be renewed. The new housing materials should fulfil the appropriate ISO standard [9] .

If the high nitrogen oxide levels are indeed due to a NOx reservoir in the dust layer, this is another argument for a total cleanup of the archive room.

However, the high I/O ratio of nitrogen oxides must be investigated. Except for the month of May, the measurements indicate that a major source of nitrogen oxides is present within the archive. It would be natural to suspect that cellulose nitrate sheets still are present within the archive. However, this is very unlikely, as all of the archive drawers were searched especially for nitrate a few years ago. Another natural source of nitrogen oxides is combustion, again in this case that is very unlikely. Instead we suggest, that what we see is a sink effect of the likely high emission of nitrogen oxides when nitrate negatives were present, now held by surface reactions in the large amount of fluffy dust on top of the cabinets. Why the indoor NO₂ and NOx concentration suddenly drops during the month of May we cannot explain - maybe the window has been opened by an archivist, thus increasing the air exchange rate.

Notes and References

[1]	ISO 18911:2000 Imaging materials - Processed safety photographic films - Storage practices. International Organization for Standardization, Geneve.
	ISO 18918:2000 Processed photographic plates - storage practice. International Organization for Standardization, Geneve.
	NBSIR 83-2795, 1983 Air quality criteria for storage of paper-based archival records, National Bureau of Standards, USA.
	BS 5454:2000 Recommendations for the storage and exhibition of archival documents. British Standard, London.

[2] TWPI is a way to express the "preservation quality" of a archive or store room. It describes the chemical decay rate for organic materials, based on the combined effect of temperature and relative humidity of the surrounding environment. The unit - years - correspond to the time it will take for a noticeably degree of chemical deterioration to occur. The concept has evolved from measurements of acid release by cellulose acetate during accelerated aging tests. In short: when evaluating storage location from climate measurements, the higher TWPI the better.
See also:
Reilly, Nishimura, Zinn (1996): New Tools for Preservation: Assessing Long-Term Environmental Effects on Library and Archives Collections. The Commission on Preservation and Access, NY, USA.

[3] Schmidt, Kemp, Bronée, Fenger (2002): "Airborne dust in a museum environment". IAP Copenhagen 2001, 4th meeting of the Indoor Air Pollution Working Group, The National Museum of Denmark, November 8th-9th 2001. http://iaq.dk/iap/iap2001/2001_20.htm

[4] Ryhl-Svendsen, Padfield, Smith, De Santi (2003): "The indoor climate in historic buildings without mechanical ventilation systems". Healthy Buildings 2003. Preprints: Singapore December 7-11, 2003.

[5] De Santis, Bellagotti, Vichi, Allegrini (2003): "Diffusive Sampling as a Tool for the Evaluation of Indoor Air Quality in the MIMIC Project". Indoor Air Quality in Museums and Historic Properties (IAQ 2003), 5th meeting of the Indoor Air Pollution Working Group, Norwich, April 28th-29th 2003. http://iaq.dk/iap/iaq2003/2003_10.htm

[6] Data from the Agency of Environmental Protection in Copenhagen: air quality station at corner of HC Andersens Boulevard and Stormgade: http://www.miljoe.kk.dk/13CF0215-B825-4AB7-BB93-D632AC1C66F0.W5Doc

[7] BVDA Environmental Gel Lifters. Art.no. B-17000. Company P-B Miljø, Enebaervej 7, DK-8850 Bjerringbro, Denmark

[8] ImageJ software: http://rsb.info.nih.gov/ij/

[9] ISO 18902:2001 Imaging materials - Processed photographic films, plates, and papers - Filing enclosures and storage containers. International Organization for Standardization, Geneve.

Acknowledgements
Jens Glastrup, National Museum of Denmark: GC/MS analysis
Ulrich Schnell, National Museum of Denmark: SEM analysis
Franco De Santis, CNR-IIA: Air pollution passive samplers and analysis

Part 2: Slides from presentation

Author to whom correspondence may be addressed:

Karen Brynjolf Pedersen
National Museum of Denmark
Department of Conservation
Brede
PO Box 260
DK-2800 Kgs. Lyngby
Denmark

E-mail: karen.brynjolf.pedersen@natmus.dk